Mu...A a . U R Na Uw Vm Mg Uu .A. Amway

Home , African fish eagle

4-“ .3; —K: mew,“

._.wa

615..

:A.

. T. mt...

umm -“ .9!

Ir.‘ u

. 8 H .

D A

A Amway:

nauwvmmguu

21...“...

. an

.A.

- .

.Iyﬁmunmwmma

A A¢M>.m..ruw.ﬂ:

. mu...

p h:

10— t

. .

313.. u

(:3: 9‘...

.W.

‘ A 1 I}!

,A Q L.»

.2. e

,uw

#fidflw mfi v _. THESIS: A

2(1) 3 ~ LIBRARY 5/7/3982sz Michigan F‘tate University

This is to certify that the

thesis entitled AFRICAN FISH EAGLES (Haliaeetus vocifer) and MARABOU STORKS (Leptoptilos crumeniferus) IN UGANDA: USE AS BIOMONITORS OF ENVIRONMENTAL CONTAMINATION

presented by

Simon Ralph Hollamby

has been accepted towards fulﬁllment of the requirements for

MASTER OF SCIENCE degree inEAIHQBIQLQGY AND DIAGNOSTIC INVESTIGATION

‘Lm, W11 6n K. Rumbeiha

Major professor

Date 04/25/03

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

PLACE IN RETURN Box to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested.

DATE DUE DATE DUE DATE DUE

A0821? 70 5 2&1?

6/01 c:/CiRC/DatoDue.pes.p.15 AFRICAN FISH EAGLES (Haliaeetus vocifer) and MARABOU STORKS (Leptoptilos crumemferus) IN UGANDA: USE AS BIOMONITORS OF ENVIRONMENTAL CONTAMINATION

Simon Ralph Hollamby

A THESIS

Submitted to Michigan State University In partial fulﬁlment of the requirements for the degree of

MASTER OF SCIENCE

Department of Pathobiology and Diagnostic Investigation

2003 ABSTRACT AFRICAN FISH EAGLES (Huiiaeetzls vocifer) and MARABOU STORKS (Leptoptilos crumem'ferus) IN UGANDA: USE AS BIOMONITORS OF ENVIRONMENTAL CONTAMINATION

Simon Ralph Hollamby

A study was designed to evaluate concentrations of persistent organic pollutants and mercury in African fish eagles (Haliaeetus vocifer), marabou storks (Leptoptilos crumeniferus) and tilapia fish (Oreochromis niloticus). in Uganda. Total mercury concentration in breast feathers; plasma concentrations of a range of persistent organic pollutants; packed cell volume and plasma chemistry values were determined for adult and nestling African fish eagles at Lake Mburo (n = 18) and Lake Victoria near Entebbe

(n =15), as well as marabou stork nestlings in Kampala (n = 21). Morphometric measurements were collected on adult ﬁsh eagles. A human and eagle food, Oreochromis niloticus were sampled for total body mercury and a range of persistent organic pollutants

(n = 18). Feather mercury concentrations were significantly (p s 0.05) lower in fish eagles at Lake Mburo than fish eagles from Entebbe and marabou storks from Kampala.

Five adult ﬁsh eagles and ﬁve Oreochromis niloticus from Entebbe had concentrations of

4,4'-DDE of less than 0.005 ppm wet weight in plasma and ﬁsh samples. The research establishes concentrations of these pollutants in these species and allows future trend analysis. African ﬁsh eagles and marabou storks meet many criteria of a suitable avian biomonitor of environmental pollution. With appropriate development, long-term research and integration with other monitoring initiatives, these species could become valuable tools to assess environmental change. Cepyright by SIMON RALPH HOLLAMBY 2003 ACKNOWLEDGMENTS

I would like to thank Wilson Rumbeiha, the major professor and committee chair for his help and advice throughout the time of this project. I was Wilson’s ﬁrst graduate student and I think we both learned greatly from the experience.

Gratitude goes to my committee members and two residency mentors, Jim Sikarskie and

Scott Fitzgerald. Jim was always there with advice and support, especially in regard to the human issues relating to wildlife research. Scott’s straight shooting approach allows students to know exactly where they stand and this student appreciated that. John

Kaneene, my ﬁnal committee member, greatly facilitated the statistical analysis for this project and helped with advice as the senior member on my committee. Thank you.

A very special thanks goes to the field workers on this research. In every way, I would not have been able to achieve anything without them. Josephine Afema filled every role possible between and during the field periods: veterinarian, interpreter, counsel, cook, guide, banker, humorist, psychologist and social worker. The title “graduate assistant” does not do her efforts justice. Perhaps this will. Thank you friend.

To Rae Gandolf for providing personality, physical and mental strength to this research.

Thanks for sharing your abundant talents. Gretchen Yurk gave me the great pleasure of watching a good student become a great veterinarian during my time on this research.

Thanks Gretchen. Amanda Norris showed me how lucky I am to do what I do but also reminded me of the world beyond wildlife. Thanks for the help Amanda. Ken Cameron

iv worked tirelessly for 5 weeks, often at considerable personal risk and with no compensation, to achieve the objectives of this research. His work ethic and lateral thinking were the difference between success and failure. Thanks Ken. Samuel Waigo yet again proved he is one of the most reliable people I have worked with. His mechanical, ﬁeld skills and work ethic were greatly appreciated. Thanks Sam.

The number of people who assisted with this project either in the USA or Uganda are too numerous to adequately acknowledge so my apologies if I have inadvertently omitted someone. Christine Dranzoa, provided the link with Makerere University and was also a good source of support throughout this project. Thanks to Terryl Grubb, Karl Strause, Al

Parker and Jerry Brandenburg who all assisted in training field personnel in tree climbing methods. Martin Okott and Malcolm Wilson both assisted in the field and provided sound ornithological advice. Bill Bowerman provided much of the framework on which the methods used in this research were based, in addition to providing field training on bald eagles. John Stuht did all the examinations for blood parasites, with the assistance of Paul

Friederich, at the Michigan Department of Natural Resources Rose Lake Pathology

Laboratory. Thanks also to the toxicology, endocrinology and clinical pathology laboratory staff at the Diagnostic Center for Population and Animal Health, Michigan

State University Veterinary Medical Center in particular, Rae Nachreiner for thyroid analysis. Kirk Stuart and Zhiqiang Yang performed the toxicological analyses. Denise

Harrison and Sherrie Lenneman helped simplify the mysteries of university administration for me. Sherrie was a source of valuable advice and humor during my time on this project. Thanks to the Uganda Wildlife Authority, especially Charles Tumwesigye and Joseph Okori; the Uganda National Council for Science and Technology. especially

Julius Ecuru and Fred Nghania; and the Uganda Wildlife Education Centre.

Thanks to British Airways, who provided subsidized airfares through the British Airways

Assisting Conservation Program and especially Maxine Kibble, who has responsibility for this program. I sincerely thank the Morris Animal Foundation, who sponsored the project on which this work is based (grant number DOIZO-78).

Finally, the biggest thanks of all goes to my parents, who have always been there for me no matter where the road has led. One could ask for better.

vi PREFACE

The chapters in this thesis are organized as independent pieces of work some of which are intended to be published in the scientiﬁc literature. As such, there may be some redundancy between chapters, especially in relation to the reporting materials and methods.

vii TABLE OF CONTENTS

LIST OF TABLES xi

LIST OF FIGURES xiii

ABBREVIATIONS xiv

CHAPTER 1: Project Summary, LiteratureReview, Hypothesis and Research Objectives. Project Summary Literature Review Mercury and avian species Persistent organic pollutants and avian species African ﬁsh eagle biology Marabou stork biology Pesticide usage in Uganda Hypothesis Research Objectives References

CHAPTER 2: Methods and Equipment Used to Sample African 23 Fish Eagle Adults and Nestlings (Haliaeetus vocifer) and Marabou Stork N estlings (Leptoptilos crumeniferus) in Uganda. Abstract 24 Introduction 25 Materials and Methods 26 Adult eagle capture 26 Marabou Stork and African Fish Eagle Nestling Capture 29 Results 30 Discussion 32 References 35

CHAPTER 3: PCV, Biochemical Values, Hematazoon Parasites and 39 Morphometric Measurements for African Fish Eagle (Haliaeetus vocifer) Nestlings and Adults at Two Sites in Uganda. Abstract 40 Introduction 41 Materials and Methods 42 Results 48 Discussion 50 References 54

viii CHAPTER 4: PCV, Biochemical Values and Survey for Hematozoon Parasites in N estling Marabou Storks (Leptoptilos crumeniferus) in Uganda. Abstract 71 Introduction 71 Materials and Methods 72 Results 77 Discussion 78 References 81

CHAPTER 5: Persistent Organic Pollutant and Mercury Concentrations in 89 African Fish Eagles (Haliaeetus vocifer), Marabou Storks (Leptoptr’los crumeniferus) and Tilapia (Oreochramis niloticus in Uganda Abstract 90 Introduction 91 Materials and Methods 95 Results 101 Discussion 103 Mercury 103 Persistent Organic Pollutants 109 References 114

CHAPTER 6: Nest Habitat Characterization of African Fish Eagles (Haliaeetus vocifer) from three sites in Uganda 128 Abstract 129 Introduction 129 Materials and Methods 130 Results 132 Discussion 134 References 137

CHAPTER 7: Assessing the potential of African Fish Eagles (Haliaeetus vocifer) and Marabou storks (Leptoptilos crumemfems) as biomonitors of environmental change. 140 Abstract 141 Introduction 141 Materials and Methods 143 Results 147 Discussion 148 Fish eagle biology and biomonitoring 148 Fish eagle diet and biomonitoring 149

ix Fish eagle reproduction and biomonitoring 151 Laboratory and ﬁeld toxicity studies: mercury 152 Feathers as a tissue for biomonitoring mercury 152 Persistent organic pollutants, ﬁsh eagles and biomonitoring 153 Fish eagle sampling methods and biomonitoring 154 Marabou storks as biomonitors 154 Conclusion 156 References 159

CHAPTER 8: Conclusion and Recommendations for Future Research 164 Conclusion 165 Recommendations 168 References 174 LIST OF TABLES

Table 3.1. Packed cell volume, plasma chemistry values and morphometric measurements from adult African ﬁsh eagles

Table 3.2. Packed cell volume and plasma chemistry values from nestling 60 African ﬁsh eagles

Table 3.3. Morphometric measurements of adult male and female African 62 ﬁsh eagles

Table 3.4. Univariable analysis of variance of morphological data by gender from African ﬁsh eagles from Lake Mburo and Lake Victoria 63 near Entebbe (n=15)

Table 3.5. Analysis of variance of plasma chemistry values in ﬁsh eagles (n = 65 33)

Table 3.6. Multivariable analysis of variance of packed cell volume in Aﬁican ﬁsh eagle (n=18) nestlings. 67

Table 4.1. Body weight, packed cell volume and plasma chemistry values from nestling marabou storks (n =20). 85

Table 4.2. Analysis of variance of plasma chemistry values in nestling marabou storks. 87

Table 5.1. Mercury concentrations in marabou storks, African ﬁsh eagles and Oreochromis niloticus. 121

Table 5.2: Results of analysis of variance of mercury concentrations in African Fish eagle feathers. 123

xi Table 5.3: Results of analysis of variance of feather mercury concentrations between Avian species from Lake Victoria. 125

Table 5.4. Results of analysis of variance of total mercury concentrations in breast feathers of marabou storks (Leptoptilos crumem'ferus) (n = 21) from Kampala, Uganda. 125

Table 5.5. Results of analysis of variance of mercury concentrations in tilapia (Oreochromis niloticus) (n = 18) from three sites in Uganda. 125

Table 6.1. Nest site characteristics for African ﬁsh eagles 139

xii LIST OF FIGURES Page

Figure 1.1. Uganda, East Africa with research sites highlighted (modiﬁed from worldatlascom) 22

Figure 2.1. Fish snare vest 38

Figure 2.2. Modiﬁed "ﬁgure of eight" slipknot used to tie nooses on the snare vest (drawings by A. R. Gandoli) 38

Figure 3.1. Technique for measuring African Fish Eagle footpad length (Modiﬁed from Bortolotti 1984 a, b) 69

Figure 3.2. Technique for measuring African ﬁsh eagle culmen length (A), bill depth (C) and. hallux length (D). (Modiﬁed from Bortolotti 1984 a, bl 69

Figure 5.1. Total mercury in breast feathers of adult and nestling African ﬁsh eagles (n = 33) 127

xiii LIST OF ABBREVIATIONS

T1222, 10, l 0- hexachloro-l ,4,4a, 5,8,8a—exahydro-endo-1,4-exo-5,8-dimethanonapthalene

AST aspartate transaminase

Chlordane 1,4,5,6,7,8,8,-octochloro—3a,4,7,7a— tetrahydro-4,7- methanoindan

Chol cholestrol

CK creatine kinase

Dieldrin 1 ,2,3,4, 10, 10- hexachloro-6,7-epoxy- 1 ,4,4a, 5,6, 7,8, 8a—octahydro—endo- 1 .4-exo-5,8- dimethanonapthalene

DDD 1,1,1 - dichloro- 2,2-bis (4-chlorophenyl) ethane

DDE 1,1,1 -— trichloro- 2,2-bis (4—chlorophenyl) ethylene

DDT 1,1,1 -— trichloro- 2,2-bis (4-chlorophenyl) ethane

DCPAH Diagnostic Center for Population and Animal Health

Endrin 1,2,3,4,10, 10- hexachloro-6,7-epoxy-1,4,4a,5,6,7,8,8a-octahydro- 1.4- endo, endo- 5,8- dimethanonapthalene

Fish Eagle Aﬁican ﬁsh eagle (Haliaeetus vocifer)

Heptachlor 1,4,5,6,7,8,8,-octochloro-3a,4,7,7a- tetrahydro-4,7- methanindane

HCH l,2,3,4,5,6- hexachlorocyclohexane

Lindane gamma isomer of HCH

MSU Michigan State University

Na/K sodium/potassium ratio

PPm parts per million

PCV packed cell volume

xiv P phosphorous

POPS persistent organic pollutants

PCBS polychlorinated biphenyls

TT4 thyroxine

TT3 triiodothyronine

TPP refrac total plasma protein measured by refractometry

TPP col total plasma protein measured by the biuret method

TT4 RIA thyroxine radioimmunoassay

TT4 ELISA thyroxine enzyme linked immonoabsorbent assay

XV Chapter 1

Project Summary, Literature Review, Hypothesis and Research

Objectives Project Summary

With a history of social and political instability, there has been little monitoring or scientiﬁc evaluation of the health of Uganda’s ecosystems (Koch RA 1996). The onset of social and political stability in the mid 19805 has led to an expanding economy, increased foreign investment and population urbanization centered in Kampala. The effects of this change on wildlife and the environment have not been adequately documented. The

Aﬁican ﬁsh eagle (Haliaeetus vocifer) is a tertiary avian predator in lake-based food chains throughout sub-Saharan Africa (Brown et a1 1982). Piscivorous raptors have proved to be valuable biomonitors of ecosystem health and environmental pollution

(Bowerman et al 2000 a). Much of the research on raptors as environmental monitors has come from countries with a developed database on species from temperate climates that

Show marked seasonality.

Marabou storks (Leptoptilos crumeniferus) are indigenous to tropical Africa where they are common to abundant in most parts of their range. In areas such as urban

Kampala, they have a cosmopolitan diet, due to their efﬁcient adaptation to a lifestyle as scavengers of human refuse (Hancock et al 1992). Their diet and exposure to pollutants in urban areas may reﬂect human exposure to the same pollutants making marabou storks potentially useful as biomonitors of toxic exposure in humans.

The hypothesis tested by this project is there is no signiﬁcant difference in persistent organic pollutant and mercury concentrations in tissues of African ﬁsh eagles and Oreochromis niloticus from Lake Victoria near Entebbe compared to those from

Lake Mburo. The objectives of this project were to quantify the concentrations of persistent organic pollutants (POPS) in the blood and feathers of populations of African ﬁsh eagles at two study sites in Uganda. One site was an urbanized area of shoreline on Lake

Victoria from Nfo Island (0°. 00' N, 32°. 26' E) to Kisubi Bay (0°. OS'N, 32°. 35’ E) near

Entebbe. The other site was Lake Mburo, a six km long freshwater lake within a 256 sq km national park Situated in South Western Uganda (0°. 39' S, 30°. 57' E). Lake Mburo is

230 km and Entebbe 40 km from the capital, Kampala. The population of Kampala is rapidly increasing and virtually contiguous with that of Entebbe (Uganda Bureau of

Statistics 2002). Concentrations of the same pollutants measured in fish eagles were quantified in whole body cross section samples of a representative eagle prey species, tilapia (Oreochromis niloticus). The whole body cross section samples were approximately 100g whole cuts including viscera, fat, Skin and bone is taken from just cranial to the dorsal fin ventrally down to just caudal to the gill arch. Oreochromis niloticus are herbivorous fish occupying the littoral zone in African lakes. As fish can be the primary source of dietary protein for human populations in this region, sampling of

fish may also give some indication of human exposure to these pollutants. Data on concentrations of these pollutants in the blood and feathers of nestling marabou storks from urban Kampala is presented for the purpose of comparison and contrast to the fish eagle data. Blood plasma was used for total polychlorinated biphenyls (PCB) and chlorinated pesticide analysis. The specific chlorinated pesticides examined were aldrin,

DDT, a-HCH, dieldrin, endrin, heptachlor and their metabolites, B-HCH, 2,4’-DDD,

4,4’-DDE, 4,4’-DDD, 2,4’-DDT, 4,4’-DDT, heptachlor epoxide and lindane and nonachlor. Breast feathers were used for total mercury analysis. Other parameters such as weight, packed cell volume (PCV), estimated age and plasma chemistry values were determined for marabou storks and African ﬁsh eagles. A survey of hematozoan parasites was performed for the marabou stork and African ﬁsh eagle populations sampled.

Certain persistent pollutants, such as the coplanar PCBs that have an antiestrogenic effect. Some PCB congeners may disrupt thyroid gland homeostasis (Yamamoto et al

1996). Therefore, plasma total thyroxine (TT4), and plasma total triiodothyronine (TT3) were also determined. The sex of the African fish eagles and marabou storks was determined by analysis of erythrocyte DNA. Characterization of nest site habitat was made and population density was estimated but not quantified. Analysis of variance was conducted to assess the association between various factors and total feather mercury, plasma chemistry parameters, and morphological characteristics of African fish eagles and marabou storks (SAS PROC ANOVA for categorical risk factors, and SAS PROC

GLM for continuous risk factors. SAS 8.2, 2001. SAS Inc. Cary, NC). These analyses were conducted both at the univariable (only one risk factor at a time) and multivariable level. Multivariable analyses were conducted to adjust the effect of selected risk factors simultaneously. The level of signiﬁcance (type 1 [a] error) was set at p = 0.05.

There has been no documentation of persistent organic pollutant or heavy metal concentrations in Aﬁican ﬁsh eagles in Uganda. A study has examined metal

concentrations (zinc, cadmium, lead, copper, iron, manganese, chromium and cobalt) in the feathers of adult marabou storks from Kampala city and surrounding areas

(Nyangababo 2003). This study did not include mercury. Only a few studies have reported plasma chemistry values for wild eagle species (Bowerman et al 2000 b; Garcia-

Montijano et al 2002). Plasma chemistries have not. been determined for wild populations of Afiican fish eagles or marabou storks. Data generated from this project will document blood plasma concentrations of POPS, feather concentrations of total mercury and help establish parameters for plasma chemistry values of Afiican fish eagles and marabou

storks at selected sites in Uganda. The difference between the data gathered from Lake

Mburo and the Lake Victoria region (that includes Entebbe and Kampala) may highlight

the effects of greater anthropogenic environmental alteration at the latter site. The data will also increase knowledge of the biology of Aﬁican ﬁsh eagles in Uganda. To this end,

morphometric measurements were determined on adult African ﬁsh eagles and

diﬁ‘erences relative to sex were compared. The project results may also provide valuable

comparisons to analogous work being conducted on tertiary avian predators in lake-based

food chains in other areas of the world, such as bald eagles (Haliaeetus Ieucocephalus) in

the Great Lakes region of North America (Bowerman et al 1993).

The project ﬁeld methods are documented, as they may prove applicable to other

raptor species or ﬁxture research on African ﬁsh eagles and marabou storks. An

assessment was made of the suitability of the development of the Aﬁican ﬁsh eagles and

marabou Stork as biomonitors of environmental health. The conclusions are presented.

Knowledge from this research should help improve the conservation management of

African ﬁsh eagles and the Ugandan lake-based ecosystems they inhabit. Literature Review

Mercury and Avian Species

Mercury can occur in a number of chemical forms in the environment. Microbial and biochemical reactions in soils and sediments can lead to the transformation of all chemical forms into methylmercury, the most toxic form (Heinz 1996). Sources of mercury can be natural or anthropogenic and include ﬂuorescent lamps, batteries, thermometers, medicines, paints, metallurgical processes, fungicides in the paper industry, fossil fuel burning and natural release such as through volcanic eruptions (Heinz

1996). Two of Aﬁica’s most active volcanoes, Nyamuragira and Nyiragongo, situated in the Democratic Republic of Congo, bordering Uganda, erupted on July 26 2002 and

January 16 2002. These eruptions were six months prior to this project’s ﬁeld periods and widespread atmospheric deposition from the eruptions occurred for months aﬁerwards

This may have contributed to the total mercury concentrations present within the lake- based ecosystems studied in this project.

Mercury is lipid soluble and bioaccumulative. It diffuses across the alveolar space into erythrocytes and the brain, accumulates in adipose tissue and can be passed into eggs. Methylmercury can have harmful effects on adult survival, reproduction, behaviour and cellular development (Burger 1994). The neurotoxic effects of methylmercury can alter nesting behavior and negatively impact reproductive success of avian species (Heinz

1996). Mercury can cause reduced egg production, lighter eggs and smaller clutches.

Hatching success and chick survival were reduced in black ducks (Anas ruprr'pes) and mallards (Anas plaryrhyncos) fed diets containing methylmercury (Finley and Stendell 1978; Heinz 1979). The effects of mercury can be somewhat antagonized by exposure to other elements such as selenium and zinc. Animals have no physiological requirement for mercury (Eisler 1987). Anthropogenic activities that may contribute to environmental mercury’concentrations include industrial pollution and use of fossil fuels that commonly occur in the industrial area along the shore of Lake Victoria near Kampala. The human population of Kampala increased from 774,241 in 1991 to a preliminary ﬁgure of

1,208,544 in 2001 and use of diesel fuel (inferior grades and quality) in Uganda increased from 125,621 in 1997 to 207,183 cubic metres in 2001 (Uganda Bureau of Statistics,

2002). Increased use of fossil ﬁrels as an energy source may be associated with increased atmospheric mercury concentrations. Through cycling, increased atmospheric mercury may lead to an increase in mercury concentrations in aquatic organisms. No studies have been reported examining the potential impact of mercury pollution on wildlife in Uganda.

Persistent Organic Pollutants (POPS) and Avian Species

Over thirty organochlorinated compounds have been used as pesticides (Blus et al

1996). The effects of 1,1,1 - trichloro - 2,2-bis (4 - chlorophenyl) ethane (DDT) on avian species have been well documented (Lincer 1975; Peakall 1993; Ratcliﬁ’e 1967). 1, 1, 1 - trichloro - 2, 2 - bis (4-chlorophenyl) ethylene (DDE), a metabolite of DDT, is known to cause adverse effects on reproductive success and eggshell thinning in wild raptor populations (Ratcliffe 1967) and in experimental studies on raptors (Lincer 1975;

Weimeyer and Porter 1970). Polychlorinated biphenyls (PCBS) have been used as insulating materials in transformers and capacitors, plasticizers in waxes, paper manufacturing, ﬂame retardants and for a variety of other industrial applications (O’Hara and Rice 1996). Various congeners of PCBS can produce teratogenic effects, such as bill defects or deformities (Gilbertson et al 1991). Some coplanar PCB congeners have been associated with dioxin like effects including wasting, thymic atrophy and endocrine and enzyme disruption in a number of species (Safe 1990). Experimentally, reproductive failure was produced with various concentrations of PCB congeners fed to chickens

(Kubiak et al 1989). Nestling bill defects in Swedish white tailed sea eagles (Haliaeetus albicilla) were believed related to PCBS (Helander 1982). In studies on wild bald eagles, productivity was signiﬁcantly and inversely correlated with concentrations of PCBS and p,p’-DDE in addled eggs (Bowerman 1993). The role of PCBS and their estrogenic/antiestrogenic reproductive effects compared to the eggshell thinning effects of DDE are still Open to debate (Bowerman, 1995, 2000 a; O’Hara and Rice 1996).

Concentrations of organochlorinated pesticides, PCBS and mercury in bald eagles in Michigan have been documented using similar methods to those in this project

(Bowerman, 1993). Organochlorinated pesticide concentrations were measured in eagle blood during 1988 to 2000, PCB and organochlorinated pesticide concentrations in eggs item 1986 to 1997 (175 eggs) and mercury in feathers from 1986 to 2000 (Bowerman et a12000 a). Similar studies on Haliaeetus Sp. have been conducted in Sweden, Siberia and the southeastern United States and are proposed for Norway (Bowerman, 2001).

Studies on POP contaminant concentrations in African ﬁsh eagles, marabou storks and other avian species in African countries are sparse. A survey of the African

fish eagle population of Lake Kariba in Zimbabwe found all eggs collected contained DDT or its metabolites (Douthwaite 1992). The study found high concentrations of mercury in adult birds and significant eggshell thinning in areas associated with high use of DDT. The author suggested that factors other than pesticide concentrations limit breeding success at the study site. An earlier study in Zimbabwe reported egg dry weight total DDT concentrations and found reduced eggshell thickness in African fish eagle eggs. (Thomson 1984). Lincer (1981) reported organochlorinated pesticide residues in a single fish eagle egg from Kenya’s Rifi Valley Lakes and Snelling et a1 (1984) reported values from a number of eggs from various sites in southern Africa. Ratcliffe’s index of shell thickness (Ratcliffe 1967) was determined for 90 Afiican fish eagle eggs collected within southern Africa. Eggshell thickness declined progressively in relation to the use of

DDT (Davies and Randall, 1989). Interestingly, one multi year study of fish eagle productivity at Lake Chivero, a polluted dam in Zimbabwe, indicated that fish eagle numbers were increasing, despite years of eutrophication and heavy metal, sewage and pesticide effluent (Mundy & Couto 2000). The authors concluded, despite large decreases in eggshell thickness, that productivity was unaffected. The reason for the increases in productivity (in the largely piscivorous fish eagle) were thought to be an increased fish ' population caused by nutrient enrichment leading to proliferation of aquatic plant life.

The authors concluded however, heavy metal concentrations may be a cause for future concern.

The lack of uniformity in relation to tissues and species sampled, differences in

analytical methods and reporting of study results can complicate interpretation of POP

data in avian species. We report all mercury and POP concentrations in part per million (ppm). This is on a dry weight basis for feathers and a wet weight basis for whole body cross section samples of ﬁsh and avian plasma samples.

African Fish Eagle Biology

Much has been recorded on the biology of the African ﬁsh eagle (Brown, 1960,

1971, 1978, 1980; Brown & Hopcraft 1973; Erikson & Skarpe 1989; Ghiglieri 1983;

Green 1964; Krueger 1997; Prout Jones & Milstein 1980, 1986; Stewart et al 1997;

Sumba 1986, 1988, 1989; Virani 2001). Despite this information, there appears to be considerable debate over some key factors of fish eagle breeding and biology, including the proximate factors that stimulate breeding at tropical latitudes. Some of the information presented in these reports is anecdotal as opposed to scientifically designed studies. However, the majority of the studies provide valuable information based on observation but involved no capture or handling of fish eagles. One study examined nestling growth in African fish eagles (Sumba 1988) in Uganda while another measured grth in captive birds (Prout-Jones & Milstein 1986). One paper examined breeding seasonality of fish eagles in Queen Elizabeth National Park, Uganda (Sumba 1986) and concluded that breeding occurred year round with peak laying during the long wet season from August to November. No breeding records are available for Lake Mburo or Lake

Victoria at Entebbe. However, population counts have been conducted usually twice yearly at Lake Mburo and Lutembe Bay, Lake Victoria (close to the Entebbe study site)

(Nature Uganda, The East Aﬁica Natural History Society, P.O.Box 27034, Kampala,

Uganda). Our limited observations would suggest that egg laying occurs mid March to late August at Lake Mburo. Chicks of various ages from hatchlings to almost fledged as

10 well as eggs were observed in nests between late June and late August. Chicks at various stages of development were noted in nests in January, March and August at Lake Victoria near Entebbe. Eggs were observed in nests around Entebbe in August, December and

January. To accurately determine the breeding season would require multi year analysis as the proximate factors that stimulate breeding are variable and not adequately documented (Virani 2001). However, it would be reasonable to assume these proximate factors could cause variation in the timing of breeding from year to year.

Four authors examined population structure and densities of fish eagles in various regions of Uganda (Brown 1970; Green 1964; Krueger 1997; Sumba 1988). African fish eagles usually build large, conspicuous nests in tall trees close to water. Breeding may not occur every year. The incubation period for fish eagle eggs is generally believed to be around 42 days. One to three (usually two) eggs are laid at two to three day intervals.

Fledglings leave the nest at approximately 70 -75 days according to Brown (1980), while

Sumba claimed the mean fledging time to be closer to 76 days (Sumba, 1988). Fish eagles attain adult plumage at five years of age. Brown (1980) recorded adult fish eagle weights as 1.98-2.49 kg for males (n = 4) and 3.00 to 3.63 kg for females (n = 3). The literature cited in this thesis, as well as our observations confirm fish eagles as one of the most territorial avian species, both to con-specifics and other avian species. Fish eagles are reported to be mainly piscivorous but birds, small mammals, reptiles, amphibians and other avian species can make up to 9 per cent by weight of the diet in some individuals

(Stewart KM. 1997). Our observations, and the communications of local ﬁsherman, suggest that Oreochromis niloticus, usually below 25 cm in length and approximately

11 400g form the majority of the ﬁsh eagle’s diet at Lake Mburo and Lake Victoria near

Entebbe. However, diet can be variable and at Murchison Falls, the carnivorous tiger ﬁsh

(Hydrocynusforsaklii) appears to be the main ﬁsh prey item.

Marabou Stork Biology

Unlike the ﬁsh eagle, the biology and reproductive cycle of the marabou stork in

Uganda has been well described (Pomeroy 1973, 1975, 1977, 1978a, 1978b). Adult marabous weigh between 5 - 8 kg and diet of the birds in Kampala probably includes almost anything organic, such as garbage, ﬁsh remains, abattoir refuse and a large amount of vegetable matter (Brown, 1982). However, during the period of nestling growth, increased amounts of protein are taken in the form of ﬁsh, frogs and rodents.

Marabous undertake short, mainly north south migrations in Uganda, coinciding with rainfall seasonality (Pomeroy, 1978a, 1978b). The Kampala population is greatest during the dry season around December and January. Marabous are colonial nesters with multiple nests being built in particular tree types, such as Mvule (Chlorophora excelsa) and Tabebuia pentaphylla. The incubation period is 30.3 days and average clutch size is

2 - 3 eggs. Marabous have an extremely long period from hatching to fledging, being about 135 days with first flights out of the nest at 110 - 115 days. Marabous first breed at

6 - 7 years and may live up to 25 years. Breeding success is low but has reportedly increased in recent years in Kampala (Hancock et a1, 1992). To the best of our knowledge, no studies have quantiﬁed POP concentrations or plasma chemistry values in wild marabou storks. A study has examined metal concentrations (zinc, cadmium, lead, copper, iron, manganese, chromium and cobalt) in the feathers of adult marabou storks

12 from Kampala city and surrounding areas (Nyangababo 2003). This study did not include mercury.

Pesticide Usage In Uganda

Accurate data on use of pesticides and monitoring of pesticide residues in Uganda are limited (Baliddawa 1991; Ejobi et al 1996 a,b; Kock 1996; Ogutu-Ohwayo 1997;

Simonich and Hites 1997; Tukahirwa 1984, 1991; Wiktelius et a1 1999). Ejobi et al

(1996) quotes 80 tonnes of DDT per year used mainly for cotton growing and mosquito control. Dieldrin was used at the annual rate of 392 tonnes per year for banana weevils and termites while 30 tonnes was also used for Tsetse ﬂy control. Lindane, aldrin, hexachlorobenzene, campheclor, chlordane and heptachlor were also used. Ofﬁcial

ﬁgures can only be viewed as estimates of present day usage. It is clear from the literature that there is a regulatory need to evaluate and standardize methods of reporting and monitoring the use of potential chemical contaminants in Uganda. Ejobi et al (1996 a,b,) measured concentrations of DDT metabolites and other chlorinated hydrocarbons in human milk and cows milk in urban Kampala and a rural district. Concentrations of

DDT in Ugandan mother’s milk was higher than that in many developed countries such as Japan, Sweden and the USA, that had banned this pesticide. However, concentrations of DDT were lower than other developing countries such as Nigeria, India, Kenya and

Ethiopia. No studies were found examining POPS or mercury in Aﬁican ﬁsh eagles in

Uganda.

13 Hypotheses

The hypothesis to be tested was:

0 There is no difference in persistent organic pollutant and mercury concentrations

in tissues of African ﬁsh eagles (Haliaeetus vocifer) and tilapia (Oreochromis

niloticus) ﬁ'om Lake Victoria near Entebbe compared to those from Lake Mburo.

Two important issues related to the hypothesis that will be examined in this thesis are:

0 Whether concentrations of persistent organic pollutants and mercury are likely to

be Signiﬁcant contributing factors to the population dynamics of the African ﬁsh

eagle at Lake Victoria near Entebbe or Lake Mburo.

0 Whether Aﬁican ﬁsh eagles and marabou storks can be utilized as successful

biomonitors of mercury and persistent organic pollutants.

Research Objectives

The objectives of this study are to determine:

1. concentrations of organochlorinated pesticides and total PCBS from African ﬁsh

eagle and marabou Stork plasma from the stated study sites.

14 . concentrations of total mercury in breast feathers ﬁ'om African ﬁsh eagles and

nestling marabou storks from the stated study sites.

. concentrations of organochloinated pesticides, total PCBS and mercury from whole

body cross section samples of Oreochromis niloticus ﬁsh from Lake Mburo, Lake

Victoria and Murchison Falls.

. packed cell volumes (PCV) and selected plasma chemistry values from African ﬁsh

eagles and marabou storks from the stated study sites.

. plasma total thyroxine (TT4) and plasma total triiodothyronine (TT3) levels from

African ﬁsh eagles and marabou storks at the stated study sites.

. sex of all birds sampled by means of DNA analysis.

. the presence and identity of blood hematazoa from ﬁsh eagle and marabou stork

samples ﬁ'om the stated study sites.

. morphometric measurements on adult ﬁsh eagles and band (ring) all adult eagles.

. nest site habitat characteristics such as nesting tree species, tree height, nest height,

diameter of tree at breast height (DBH), disturbance around nest, distance ﬁom the

nest to water and canopy cover.

15 References

Baliddawa CW 1991. Agricultural extension and pest control: a case study of Iganga district. Insect. Sci. Applic. 12: 579-583.

Blus LJ, Weimeyer SN & Henny CJ. 1996. Organochlorine Pesticides. In: A F airbrother, LN Locke & GN Hoff, (Eds). Noninfectious Diseases of Wildlife. Iowa State University Press. Ames, Iowa. pp. 61-70.

Bowerman WW. 1993. Regulation of bald eagle (Haliaeetus leucocephalus) productivity in the great lakes basin: an ecological and toxicological approach. PhD Thesis. Michigan State University.

Bowerman WW, Geisey JP, Best DA & Kramer VJ. 1995. A review of factors affecting productivity of bald eagles in the great lakes region: implications for recovery. Environ. Health Persp. 103: 51-59.

Bowerman WW, Best DA, Grubb TG, Sikarskie JG & Giesy JP. 2000 a. Assessment of environmental endocrine disruptors in bald eagles of the great lakes. Chemosphere 41: 1569-1574.

Bowerman WW, Stickle JE, Sikarskie JG & Giesy JP. 2000 b. Hematology and serum chemistries of nestling bald eagles (Haliaeems leucocephalus) in the lower peninsula of Michigan, USA. Chemosphere 41: 1575-1579. 2000.

Bowerman WW. Pers. Corn. Feb. 16, 2001. Researched bald eagle and other Haliaeetus sp. toxicology for 16 years. Assistant Professor of Wildlife Toxicology, Clemson University, North Carolina.

Brown L. 1960. The Aﬁican Fish Eagle Haliaeetus vocifer especially in the Kavirondo Gulf. Ibis. 102: 285-297.

Brown L. 1971. Aﬁican Birds of Prey. Houghton Mifﬂin Company. Boston.

Brown L. 1978. Sexing the Fish Eagle. Bokmakierie. 30: 108.

16 Brown L. 1980. The Aﬁican Fish Eagle. Bailey Bros. & Swinfen Ltd., Folkestone, England.

Brown L, Urban EK & Newman K. 1982. The Birds of Africa. 1. Academic Press, New York. p. 188.

Brown LH & Hopcrafi JBD. 1973. Population structure and dynamics in the African fish eagle Haliaeetus vocifer (Daudin) at Lake Naivasha, Kenya. E. Afiic. Wildl. J. 11: 255- 269.

Burger J. 1994. Metals in avian feathers: bioindicators of environmental pollution. Rev. Environ. Toxicol. 5: 203-31 1.

Davies RAG & Randall RM. 1989. Historical and geographical patterns in eggshell thickness of African ﬁsh eagles Haliaeetus vocifer, in relation to pesticide use within southern Africa. In: Raptors in the modern world. BU Meyburg & Chancellor RD (Eds) World Working Group on Birds of Prey and Owls. Berlin, London & Paris. 1-611. 501- 513.

Douthwaite RJ. 1992. Effects of DDT on the ﬁsh eagle Haliaeetus vocifer population of Lake Kariba in Zimbabwe. Ibis. 134:250-258.

Ejobi F, Kanja LW, Kyule MN, Muller P, Kruger J. & Latigo AAR. 1996a. Organochlorine pesticide residues in mothers’ milk in Uganda. Bull. Environ. Contam. Toxicol. 56: 873-880. 1996 a.

Ejobi F, Kanja LW, Kyule MN, Muller P, nger J, Nyeko JPH & Latigo AAR. 1996b. Organochlorine Pesticide Residues in Cow’s Milk in Uganda. Bull. Environ. Contam. Toxicol. 56: 551-557.

Eisler R. 1987. Mercury hazards to ﬁsh, wildlife and invertebrates: a synoptic review. US. Fish and Wild. Serv. Biol. Rep. 85: 90-110.

Eriksson MOG & Skarpe C. 1989. Density and distribution of the African ﬁsh eagle and the osprey at Lake Kariba. Honeyguide. 35:54-62.

17 Finley MT & Stendell RC. 1978. Survival and reproductive success of black ducks fed methymercury. Environ. Poll. 16: 51-64.

Garcia-Montijano M A, Lemus J A, Montesinos A, Canales R, Luaces l & Pereira P. 2002. Blood chemistry, protein electrophoresis, and hematological values of captive Spanish imperial eagles (Aquila adalberti). J. Zoo Wildl. Med. 33: 112-117.

Ghiglieri MP. 1983. A riverside count in Africa’s Selous game reserve. Oryx 17: 26-27.

Gilberstson M, Kubiak T, Ludwig J. & Fox G. 1991. Great lakes embryo mortality, edema and deformities syndrome (GLEMEDS) in colonial ﬁsh eating birds: similarity to chicken edema disease. J. Toxicol. Environ. Health. 33:455-520.

Green J. 1964. The numbers and distribution of the Aﬁican ﬁsh eagle Haliaeetus vocifer on the eastern shores of Lake Albert. Ibis. 106: 125-128.

Hancock JA, Kushlan JA & Kahl PM. 1992. Storks, Ibises and Spoonbills of the World. Academic Press Inc. San Diego, California. pp. 133-138.

Heinz GH. 1979. Methylmercury; reproductive and behavioural effects on 3 generations of mallard ducks. J. Wildl. Manage. 43: 394-401.

Heinz GH. 1996. Mercury poisoning in wildlife. In: A Fairbrother, LN Locke & GN Hoff, (Eds). Noninfectious Diseases of Wildlife. Iowa State University Press. Ames, Iowa Pp. 188-127.

Helander B, Olsson M & Reutergardh L. 1982. Residue levels of organochlorine and mercury compounds in unhatched eggs and the relationship to breeding success in white tailed sea eagles Haliaeetus albicella in Sweden. Holarctic Ecol. 5: 349-366.

Kock. RA 1996. Needs Assessment for Wildlife Veterinary Activities in Uganda. Ugandan Wildlife Authority. Ministry of Tourism, Wildlife & Antiquities. p 10.

Krueger O. 1997. Population density and intra- and interspecific competition of the Afiican fish eagle Haliaeetus vocifer in Kyambura Game Reserve, southwest Uganda. Ibis 139:19-24.

l8 Kubiak TJ, Harris H], Smith LM, Scwartz TR, Stalling DL, Trick JA, Sileo L, Docherty DE & Erdman TC. 1989. Microcontaminants and reproductive impairment of the Foster’s tern on Green Bay, Lake Michigan: 1983. Arch. Environ. Contam. Toxicol. 18: 706-727.

Lincer J. 1975. DDE induced eggshell thinning in the American kestrel; a comparison of the ﬁeld situation and the laboratory results. J. Applied Ecology 12:16-21.

Lincer JL, Zalkind D, Brown LH & Hopcraft J. 1981. Organochlorine residues in Kenya’s rift valley lakes. J.Appl. Ecol. 18: 157-171.

Mundy P J & Couto J T. 2000. High productivity by ﬁsh eagles on a polluted dam near Harare. Ostrich. 71: 11-14.

Nyagababo JT. 2003. Trace metals in the contour feathers of marabou stork (Leptoptilos’ crumeniferus) ﬁ'om Kampala city and its surrounding areas. Bull. Environ. Contam. Toxicol. 70: 792-799.

Ogutu-Ohwayo R. 1997. Human Impacts on the African Great Lakes. Env. Biol. of Fishes 50: 117-131.

O’Hara TM & Rice CD. 1996. Polychlorinated biphenyls. In: A Fairbrother, LN Locke & GN Hoff, (Eds). Noninfectious Diseases of Wildlife. Iowa State University Press. Ames, Iowa pp. 71-86.

Peakall D. 1993. DDE induced eggshell thinning: an environmental detective story. Env. Rev. 1: 13-20.

Pomeroy DE. 1973. The Distribution and Abundance of Marabou Storks in Uganda. E. Aﬁic. Wildl. J. 11: 227-240.

Pomeroy DE. 1975. Birds as Scavengers of Refuse in Uganda. Ibis. 117: 69-81.

19 Pomeroy DE. 1977. The biology of marabou storks in Uganda. 1. some characteristics of the species, and the population structure. Ardea. 65: 1-24.

Pomeroy DE. 1978a. The biology of marabou storks in Uganda. II. breeding biology and general review. Ardea. 66: 1-23.

Pomeroy DE. 1978b. Seasonality of marabou storks Leptoptilos crumeniferus in Eastern Africa. Ibis. 120: 313-321.

Prout-Jones DV & Milstein PLS. 1980. Field sexing in the ﬁsh eagle. Bokmakierie 32:78-84.

Prout-Jones DV & Milstein PLS. 1986. Sequential molt with age class establishment in the Aﬁican ﬁsh eagle (Haliaeetus vocifer). S. Afr. J Wildl. Res. 16:17-26.

Ratcliffe DA 1967. Decrease in eggshell weight of certain birds of prey. Nature 215: 208-210.

Safe S. 1990. Polychlorinated biphenyls (PCBS), dibenzo-p-dioxins (PCDDs), dibenzofurans (PCDFS), and related compounds: environmental and mechanistic considerations which support the development of toxic equivalency factors (TEFS). Crit. Rev. Toxicol. 21: 51-88.

Simonich SL & Hites RA. 1997. Relationships between socioeconomic indicators and concentrations of organochlorine pesticides in tree bark. Envir. Sci Tech. 31: 999-1003.

Snelling JC, Kemp AC & Lincer JL. 1984. Organochlorine residues in Southern African raptor eggs. In: Proceedings of the Second Symposium on Africa Predatory Birds. Natal Bird Club, Durban. pp. 161-168.

Stewart KM, Matthiesen DP, Leblanc, L. & West, J. 1997. Prey diversity and selectivity of the African ﬁsh eagle: data ﬁom a roost in northern Kenya. Afr. J. of Ecol. 35: 133- 145.

Sumba SJA 1986. Breeding seasonality of the African ﬁsh eagle in Queen Elizabeth Park, Uganda. Afr. J. of Ecol. 24:103-110.

20 Sumba SJA. 1988. Nestling growth in the African ﬁsh eagle in Uganda. Afr. J. of Ecol. 26: 315-321.

Sumba SJA. 1989. Food procurement through piracy and scavenging in the African ﬁsh eagle in Queen Elizabeth National Park, Uganda. Afr. J. of Ecol. 27: 111-118.

Thomson WR. 1984. DDT in Zimbabwe. In: Proceedings of the Second Symposium on African Predatory Birds. Natal Bird Club.

Tukahirwa EM. 1991 Integrated pest management: options for Uganda. Insect. Sci. Applic. 12: 535-539.

Uganda Bureau of Statistics. 2002. Key Economic Indicators. 47‘” Issue. Ch 39. Uganda - Bureau of Statistics, Entebbe, Uganda.

Virani M. Field Omithologist, the Peregrine Fund, Kenya. E mail communication. February, 2001

Wiemeyer SN & Porter RD. 1970. DDE thins eggshells of captive American kestrels. Nature. 227:737-738.

Wiktelius S, Chiverton PA, Meguenni H, Bennaceur M, Ghezal F, Umeh EDN, nguatu RI, Minja E, Makusi R, Tukahirwa E, Tinzaara W & Deedat Y. 1999. Effects of insecticides on non-target organisms in African agroecosystems: a case for establishing regional testing programmes. Agric. Ecosys. Env. 75: 121-131.

Yamamoto JT, Donohoe RM, Fry DM, Golub MS & McDonald JM. 1996. Environmental estrogens. In: A Fairbrother, LN Locke & GN Hoff, (Eds). Noninfectious Diseases of Wildlife. Iowa State University Press. Ames, Iowa. Pp. 31-51.

21 Figure 1.1. Uganda, East Africa with research sites highlighted (modiﬁed from worldatlas.com)

Murchison Falls National Park

“FMMTZT ﬂ#.4 ‘ ; o ,4th 1 co i

Entebbe/Kampala: Lake Victoria

22 Chapter 2

Methods and Equipment Used to Sample African Fish Eagles

(Haliaeetus vocifer) and Marabou Storks (Leptoptilos crumemjferus) in

Uganda

23 Abstract

A study was designed to evaluate persistent organic pollutants and mercury levels in African ﬁsh eagles (Haliaeetus vocifer), marabou storks (Leptoptilos crumeniferus) and tilapia (Oreochromis niloticus) in Uganda. The objective of this paper is to describe the methods and equipment used in the study. Birds were sampled from the

Kampala/Entebbe region and Lake Mburo in Uganda, in three ﬁeld periods from

December 2001 through January 2003. Adult eagles were captured on water using a ﬁsh

“snare vest”. The ratio of number of birds caught to eagle attempts to take the snared ﬁsh was 1: 6 at Lake Mburo and 1:10 at Lake Victoria near Entebbe. The ratio of the number of birds caught to the number of times the snared ﬁsh was placed in the water was 1:8 for

Lake Mburo and 1:36 for Lake Victoria. One hundred and twenty capture attempts were made over ten days at Lake Mburo and 72 over nine days at Lake Victoria near Entebbe.

Eighty-three percent of adult eagles were snared by the third digit. Nestling marabou storks and African fish eagles were captured using professional tree climbing methods and equipment. The sampling success rate, defined as trees climbed with at least one chick sampled was 100% for storks and 55% for eagles. The snare vesting technique described may be an effective method to catch adult Afiican fish eagles with a success rate dependant on multi-factorial local site conditions. The climbing methods described were successful for the safe sampling of nestling marabou storks and Afiican fish eagles.

These capture methods may prove usefiil in the management and study of captive and wild populations of African fish eagles and marabou storks. Snare vesting, with modifications, knowledge of local site conditions and species biology could have applicability to the capture of other large piscivorous eagle species.

24 Introduction

The African ﬁsh eagle (Haliaeetus vocifer) has a wide distribution along lakes and waterways throughout sub-Saharan Aﬁica (Brown, 1980). A number of authors have described the biology of the species in various locations (Brown 1960, 1971, 1980;

Brown and Hopcraft 1973; Green 1964; Krueger 1997; Sumba 1986, 1988, Prout Jones &

Milstein 1986). These studies predominantly involved non-invasive observations.

Studies examining the effects of persistent organic pollutants (principally DDT and its metabolites) on reproductive success in African ﬁsh eagles were conducted from the

1970s through the 19905 and mainly in southern Aﬁica (Douthwaite 1992; Davies and

Randall 1989; Lincer 1981; Snelling et a1 1983; Thomson 1984). These studies examined eggs. However, the method used to obtain the samples are often only brieﬂy described.

The use of blood for pesticide analysis is advantageous as the effect on the population is less than the permanent removal of eggs for analysis. Methods have been described in certain species whereby total DDT residues in plasma can be adjusted to estimate the residues in the egg. (Henny and Meeker 1981). However, the use of blood as a sample tissue requires catching the bird. No known studies have described in detail the temporary capture of adult and nestling Afiican fish eagles for scientific research.

The biology and breeding cycle of marabou storks (Leptoptilos crumemferus) in

Uganda has been well described (Pomeroy 1977, 1978). One report described the use of oral anesthetic agents for capture and sampling of adult marabou storks, but reported some degree of mortality associated with the method (Pomeroy and Woodford 1976). No known studies describe in detail the temporary capture of nestling marabou storks for

25 blood collection. With the increasing scrutiny all types of invasive wildlife field investigations are currently subjected to, and to facilitate fiiture research, it behooves all researchers to adequately document and disseminate successfirl capture methods. It is also important to highlight unsuccessful capture methods so other researchers can assess, then modify, improve on, or discard the techniques. This is vital because as populations of many Species continue to decline, proven and safe capture methods are required. This

report describes methods used to capture adult and nestling Aﬁican ﬁsh eagles and

nestling marabou storks in the Kampala/Entebbe region and at Lake Mburo National Park

in Uganda, East Aﬁica during three ﬁeld periods from December 2001 through January

2003. The captures formed part of a study to collect feathers for mercury and plasma for

detemrination of organochlorinated pesticides, total PCBs, and plasma chemistry values in African ﬁsh eagles and marabou storks. In addition, morphometric data was collected.

Materials and Methods

Adult Eagle Capture

Aﬁican ﬁsh eagles were sampled at Lake Mburo, a six km2 freshwater lake in

South Western Uganda (0°. 39' S, 30°. 57' E) situated in a 256 km2 national park, and on

Lake Victoria at Entebbe, from Nfo Island (0°. 00' N, 32°. 26' E) to Kisubi Bay (0°. 05' N,

32°. 35' E). Fish eagles were sampled in July, August and December 2002 at Lake Mburo

and August 2002 and January 2003 at Entebbe. Forty ﬁve percent of birds were sampled

between 0600 and 1200 hours and 55% between 1200 to 1800 hours.

26 Adult ﬁsh eagles were captured on the water using a “snare vest” technique. At Lake

Mburo, the boat used was an inﬂatable 3 m x 1.46 m Seaeagle 8 with motorrnount

(SeaEagle Boats, Port Jefferson, NY) crewed by three and powered by a ﬁve horsepower

(hp) Yamaha outboard motor (Yamaha Motor Corporation, Kennesaw, GA). At Entebbe, a 6.1 m local fishing vessel crewed by four and powered by Yamaha outboards, ranging from 5 to 40 hp was used. Fish fitted with the snares were tilapia (Oreochromis niloticus) between 15-25 cm in length and approximately 300 - 400 g. A ventral midline incision from the anogenital opening to the lateral fins was made and the viscera removed. Foam and small pebbles were inserted in the coelomic cavity and mouth so the fish would float laterally on the water surface. The body cavity and mouth were sewn closed with a simple continuous suture pattern using 8 pound test (3.6 kg) 0.12” diameter monofilament fishing line (Stren Fishing Lines, Madison, NC). Multiple loops used to make the snares were constructed from 25 pound test (11.3 kg), clear or pale green, monofilament nylon, 4.8 mm (0.19”) diameter fishing line (Pure Fishing, Spirit Lake,

IA). To create the snares, a 60-80 cm piece of 25 pound test line was cut and a “ﬁgure of eight” knot tied loosely in one end (Figure 2.2 a,b). The end of the line used for the ﬁnal knot was then placed through one circle of the eight to create a slipknot and excess line was pulled thru to make 5-6 cm diameter snares and a free end of line (Figure 2.2 c).

Hemostats and tension were then applied to the line to loosely set the knot and excess line near the slipknot was cut. Eight to twelve snares were made per ﬁsh. The free end of the line was threaded through an eyed needle that was used to penetrate the body of the ﬁsh.

This allowed the line to exit on the underside of the fish as it floated laterally in the water. Most often, snares were placed at equidistant intervals to cover the whole floating

27 surface of the fish. The free ends of the line were brought together and held in place by a lightly clamped hemostat. They were then tied on themselves with multiple square knots and the excess line cut. The completed snare vest (Figure 2.1) was then set aside until just before use. Sixty pound test (27 kg), 6.3 mm (0.25”) diameter, monofilament fishing line

(Danielson Company Inc., Auburn, WA) was used to attach the snare vest to a wooden reel, manually held by an operator after being attached to the boat. To prepare the snare vest for immediate use, the free ends of the 25 pound test line were attached to the 60 pound test line by a series of square knots. The snared fish was then placed in the water and allowed to gain distance fi'om the boat by passively drifting or actively paddling away from the fish. Distance from the boat ranged from 6.1 m to 30 m. Once an eagle was entangled by a digit in a snare, the field crew paddled to it while maintaining tension on the line. Shoulder length, kevlar lined animal handling gloves (VetPro Warden gloves,

Medical Service Associates, Newington, CN) were used to retrieve the eagle by securing both legs in the tibiotarsal region. A second operator then covered the head with a bag and supported the body until the bird could be taken to shore for sampling. Two pairs of gloves were used to ensure safe transfer of the bird from the operator in the boat to an operator on shore. Fish eagles swim well so there was little risk of drowning. Once all samples had been collected and measurements taken and recorded, the eagle was placed in a large cotton sack, weighed and released from land at the closest point possible to the

capture location. Average time from capture to release was 32 minutes (range 20-45).

28 Marabou Stork and African Fish Eagle Nestling Capture

Marabou stork nestlings were sampled in January 2003 in the center of Kampala city, along Nile Avenue, one of the main thoroughfares. All nests were in the introduced white cedar tree (T abebuia pentaphylla). Fish eagle nestlings were sampled at Lake

Mburo and Entebbe locations as previously described for adults. Marabou Stork and ﬁsh eagle nestlings were temporarily captured for sample collection using professional tree climbing methods based on those described in the National Tree Climbing Field Guide

(USDA Forest Service 1996), with modiﬁcations for tropical tree species and environmental conditions. The majority of the trees climbed were Chlorophora excelsa,

Antiarus toxocara and Acacia sieberiana. Average tree height was 29.87 m at Entebbe

(range 17.10 - 46.33 m) and 11.58 m at Lake Mburo (range 6.40 - 18.89 m) and average nest height was 22.55 m (range 12.80 - 34.74) at Entebbe and 8.84 m (range 3.96 - 16.46) at Lake Mburo. The main method of tree ascent was using tree climbers with 70 mm gaffs (Klein Tools Inc., Chicago, IL), a leather climbing saddle (Weaver Leather Inc, Mt.

Hope, OH), locking carabiners (Petzl America, Clearﬁeld, UT), tubular webbing and lanyards. Lanyard types used were 13 mm steel core 3.7 m (New England Ropes, Fall

River, MA) combined with microcenders (Petzl America, Clearﬁeld, UT) or a two in one

Prusik lanyard. Rope type used for rappelling and occasional ascending was Kemmantle

46 m, 11mm diameter static line (New England Ropes, Fall River, MA).

When necessary ﬁsh eagle and stork nestlings were gently coaxed to the side of the nest using an “eagle hook”. This was either a converted telescopic car aerial (Scosche, telescOpic fender mount antennae, Oxnard, CA) twisted at the end to form a hook or an ice gaff (Mason Tackle Company, Otisville, MI), with the point of the hook covered with

29 protective plastic. The ice gaff was also used by the climber to fend off Marabou storks,

who often were very protective of their young. Adult ﬁsh eagles did not interfere with

sampling of nestlings. Fish eagle and stork nestlings were placed singly into a 40 cm

diameter nylon bag speciﬁcally designed for similar work with bald eagle (Haliaeerus

Ieucocephalus) nestlings (The Taku Tailor, Juneau, AK). The bag had a padded bottom,

velcro tab fasteners and multiple ventilation holes. The bag was attached to a rope and

lowered to the ground for sampling. Average time from capture to release for ﬁsh eagle

nestlings was 32 minutes (range 10-60 minutes) and for marabou stork nestlings was 15

minutes (range 7-22 minutes).

All procedures utilizing birds in this study were carried out under approval of the

Michigan State University All University Committee on Animal Use and Care. Research

permits were granted by the Uganda National Council for Science and Technology and

the Uganda Wildlife Authority.

Results

Twelve adult eagles were captured (ten at Lake Mburo and two at Entebbe) using

the snare vest method. The ratio of number of birds caught to attempts eagles made to

take the snared ﬁsh was 1: 6 at Lake Mburo and 1:10 for Lake Victoria at Entebbe. The

ratio of the number of birds caught to the number of times the snared ﬁsh was offered to a

ﬁsh eagle, irrespective of whether an attempt was made to catch the snared ﬁsh was 1:8

' for Lake Mburo and 1:36 for Lake Victoria. One hundred and twenty capture attempts

were made over ten days at Lake Mburo and 72 attempts over nine days at Entebbe. Ten

30 eagles (83%) were trapped by a snare encircling the bird’s third (middle) digit. There appeared to be no consistent area of the ﬁsh that was struck by the birds.

Eighteen ﬁsh eagle nestlings were sampled by the tree climbing methods described. Eight nestlings were sampled at Lake Mburo from ﬁve nest sites and ten from

Lake Victoria at five nest sites. Nestlings ofien could not be visualized, even with binoculars before trees were climbed. At two nests (containing two nestlings each) the nestlings were at fledging age and flew off when the climber approached the nest. Two returned to the nest unaided while two swam to the lakeshore where they were retrieved, sampled and returned to the nest. Nests were viewed daily for an average period of three days afier sampling and no problems such as abandonment or interrupted feeding were noted. However, visualization of the chicks was often impossible from the ground.

Parents returned to their chicks either immediately after the retreat of the climber, or soon thereafter.

Twenty-one marabou stork nestlings were sampled from twelve nests, in six trees in one colony during an eleven-hour period. Parents either never left the nest (and vigorously defended the chicks) or remained close to the nest and returned upon retreat of the climber. The success rate, deﬁned as trees climbed with at least one chick sampled

(nestlings under 800g were deemed too small to sample safely) was 100% for storks and

55% for eagles.

31 Discussion

The success of the snare vesting capture technique for African ﬁsh eagles was Site dependant. The reasons that snare vesting was more successful at Lake Mburo than Lake

Victoria were multi-factorial. Fish eagle population density is greater per 100 meters of shoreline at Lake Mburo than Entebbe. The density allowed the researcher’s to exploit the extreme territoriality of fish eagles to aid the capture process. Laying a snared fish on the suspected borderline between two territories created competition for the food resource and led to more attempts to take the fish. Competition occasionally led to a rushed, less calculated approach to the snared fish by the eagle thus increasing capture success.

Smaller territory size and closer placement of the snared fish to the eagle perching trees also appeared to increase the number of eagles caught at Lake Mburo. The smaller inflatable boat used at Lake Mburo was more maneuverable than the large wooden boat used on Lake Victoria. This facilitated quick and accurate placement of the snared fish at a desirable site. Quick placement of the snared fish was important as many eagle attempts to take the snared fish occurred five minutes or less after snare placement. Furthermore, the usually calmer waters at Lake Mburo helped with snare placement and thus capture success. Optimum time of day for capture at both sites was early morning (0700-0900 hr) and late afternoon (1600-1800 hr), corresponding to observed peak eagle fishing activity.

Color and size of fishing line was important, as eagles would avoid fluorescent green line and “abort” their approach to the snare at the last second. Clear line was the ideal color. Lines heavier than 25-pound test (11.3 kg) became too difficult to create small easily knotted nooses.

32 When releasing eagles, it was important to choose the release site carefully. The release site had to be within the eagle’s territory or an area where the bird would not be attacked by con-speciﬁcs. The latter occasionally happened. (with no untoward sequelae) to a bird snared in the water before retrieval could occur. A release site with a view of the lake, and far enough back from the shore so the eagle could orientate itself before attempting ﬂight proved optimal. The only complication arising from this snare vesting technique was that on two occasions, an eagle broke free of the vest with a snare encircled around the third digit. One of these eagles was captured on a subsequent second attempt and the snare was no longer present. It is suspected that removal of the slipknots by the eagles could be accomplished with ease, however this was not proven.

The major variables with the snare vest technique that negatively impacted capture rates were snare number and size. Having small numbers (less than ﬁve) of larger snares (greater than 8 cm diameter) rarely proved successful. Eight to twelve snares per

ﬁsh was optimal. However, the use of four snares proved successﬁil for the capture of bald eagles (Bowerman 2001).

The high success rate for sampling marabou Stork nestlings was largely due to easy nest accessibility, the high degree of visibility of chicks from the ground and the well deﬁned seasonality of their breeding cycle (Pomeroy, 1977, 1978a). For eagle nestlings, the success rate was aided by predetermination of nest activity status, based on female eagle activity, nest size, state of nest repair and amount of droppings/prey remains

33 under the nest. The lower success rate with eagle nestlings was partially due to the lack of a deﬁned and studied breeding pattern at both our sample sites and an inability to see into the nest from the ground.

We conclude that the climbing methods described in this paper are successful for the Safe sampling of nestling marabou storks and African ﬁsh eagles. In addition, snare vesting, as described in this communication may be an effective method to catch adult

African ﬁsh eagles with a success rate that depends on multi-factorial local site conditions. The technique, with modiﬁcations, knowledge of local site conditions and species biology could have applicability to the capture of other large piscivorous eagle species.

34 References

Bowerman WW. Pers. Com. Feb.16, 2001. Researched bald eagle and other Haliaeetus sp. toxicology for 16 years. Assistant Professor of Wildlife Toxicology, Clemson University, North Carolina.

Brown LH. 1960. The African ﬁsh eagle Haliaeetus vocifer especially in the Kavirondo gulf. Ibis. 102: 285-297.

Brown L. 1971. African Birds of Prey..Houghton Mifﬂin Company. Boston.

Brown L. 1980. The African Fish Eagle. Bailey Bros. & Swinfen Ltd., Folkestone, England.

Brown LH & Hopcraft JBD. 1973. Population structure and dynamics in the African Fish Eagle Haliaeetus vocifer (Daudin) at Lake Naivasha, Kenya. E. Aﬁic. Wildl. J. 11: 255- 269.

Davies RAG & Randall RM. 1989. Historical and geographical patterns in eggshell thickness of African ﬁsh eagles Haliaeetus vocifer, in relation to pesticide use within southern Africa. In BU Meyburg & RD Chancellor ((Eds.)). Raptors in the modern world. World Working Group on Birds of Prey and Owls, Berlin, London & Paris. 501- 513.

Douthwaite RJ. 1992. Effects of DDT on the Fish Eagle Haliaeetus vocifer population of Lake Kariba in Zimbabwe. Ibis 134:250-258.

Green J. 1964. The numbers and distribution of the South African ﬁsh eagle Haliaeerus vocifer on the eastern shores of Lake Albert. Ibis. 106: 125-128.

Henny CJ & Meeker DL. 1981. An evaluation of blood plasma for monitoring DDE in birds of prey. Environ. Poll. 25: 291-304.

Krueger O. 1997. Population density and intra- and interspeciﬁc competition of the African ﬁsh eagle Haliaeetus vocifer in Kyambura game reserve, southwest Uganda. Ibis 139219-24.

Lincer JL, Zalkind D, Brown LH & Hopcraft J. 1981. Organochlorine residues in Kenya’s riﬁ valley lakes. J.Appl. Ecol. 18: 157-171.

35 Pomeroy DE & Woodford MH. 1976. Drug immobilization of marabou storks. J. Wildl. Manage. 40: 177-179.

Pomeroy DE. 1977. The biology of marabou storks in Uganda. 1. Some characteristics of the species, and the population structure. Ardea. 65: 1-24.

Pomeroy DE. 1978. The biology of marabou storks in Uganda. II. Breeding biology and general review. Ardea. 66: 1-23.

Prout-Jones DV & Milstein PLS. 1986. Sequential moult with age class establishment in the African ﬁsh eagle (Haliaeetus vocifer). S. Aﬁ'ic. J. Wildl. Res. 16: 17-26.

Snelling JC, Kemp AC & Lincer JL. 1984. Organochlorine residues in southern African raptor eggs. In: Proceedings of the Second Symposium on Aﬁica Predatory Birds. Natal Bird Club, Durban. pp. 161-168.

Sumba SJA. 1986. Breeding seasonality of the Afiican fish eagle in Queen Elizabeth Park, Uganda. Afiic. J. Ecol. 242103-110.

Sumba SJA. 1988. Nestling growth in the Aﬁican ﬁsh eagle in Uganda. Afric. J. Ecol. 26: 315-321.

Thomson WR. 1984. DDT in Zimbabwe. In: Proceedings of the Second Symposium on Aﬁican Predatory Birds. Natal Bird Club, Durban.

United States Department of Agriculture Forestry Service. National Tree Climbing Field Guide, USDA Forest Service 1996, Missoula Technology and Development Center, Missoula, Montana, 1996.

36 Figure 2.1: Fish snare vest utilizing tilapia (Oreochromis niloticus). Note fish floats laterally due to Styrofoam in the coelomic cavity and small pebbles in the mouth. A total of eight to twelve snares were made per fish.

Figure 2.2: A modiﬁed "ﬁgure of eight" slipknot was used to tie nooses on the snare vest

To create the snares, a 60-80 cm piece of 25 pound test line was cut and a “ﬁgure of eight” knot tied loosely in one end (Figure 2.2 a,b top view). The end of the line used for the ﬁnal knot was then placed through one circle of the eight to create a slipknot and excess line was pulled thru to make 5-6 cm diameter snares and a free end of line (Figure

2.2 c bottom view) (drawings by A. R Gandolf).

Figure. 2.2: Modiﬁed "ﬁgure of eight" slipknot used to tie nooses on the snare vest (drawings by A. R. Gandolf)

38 Chapter 3

Packed Cell Volume, Biochemical Values, Blood Parasites and

Morphometric Measurements for African Fish Eagle (Haliaeetus

vocifer) Nestlings and Adults at Two Sites in Uganda

39 Abstract

Packed cell volumes (PCV) and plasma chemistry parameters were measured in

15 adult and 18 nestling Afiican fish eagles (Haliaeetus vocifer) fi'om June 2002 through

January 2003. Morphometric measurements were taken on 15 adult eagles. All eagles were sampled for blood parasites and sexed by erythrocyte DNA extraction. Ten adults and eight nestlings were sampled from Lake Mburo and ﬁve adults and ten nestlings from

Lake Victoria near Entebbe in Uganda. Analysis of variance was conducted to assess the association between site, age, gender and plasma chemistry parameters and the association between gender and morphological characteristics. Plasma chemistry values reported for nestling and adult African ﬁsh eagles are similar to those reported for other captive and free-ranging eagle species. Packed cell volumes for nestlings were markedly lower than values reported for similarly aged nestlings of other eagle species. There was no signiﬁcant difference (p 2 0.05) in PCV of nestling eagles of different body weights.

There was variation in all measured plasma chemistry parameters between adults and nestlings, most significantly (p s 0.05) PCV, calcium, phosphorous, potassium, cholesterol and creatine kinase (CK), all of which were lower in adults, except aspartate transaminase (AST), which was higher. Plasmodium cicumflexium like parasites were present in the erythrocytes of three nestlings fiom Lake Mburo. Like other Haliaeetus sp. body weight, bill depth, culmen, toepad and hallux length as well as bill depth measurements were significantly (p s 0.05) greater for females than males. The data provides baseline biological and physiological information that may prove useful in the management and study of captive and wild populations of African fish eagles.

4o Introduction

The African ﬁsh eagle (Haliaeetus vocifer) inhabits lakes and waterways throughout sub-Saharan Africa (Brown, 1980). It does not appear in the appendices of the

Convention In Trade of Endangered Species (CITES). The African fish eagle fills a position of tertiary avian predator, its diet being predominantly fish (Stewart et al 1997).

This is a similar niche to that ﬁlled by the bald eagle (Haliaeetus Ieucocephalus) in lake- based ecosystems in North America. The bald eagle has been proposed as an ecosystem

monitor species of Great Lakes water quality, particularly in regard to the toxic effects of organochlorinated compounds on piscivorous wildlife (Bowerman, 2003). Given the widespread distribution of the African ﬁsh eagle and its relative abundance, it may also be a valuable indicator species of water quality in African lake-based ecosystems, such as

Lake Victoria. However, there are many factors that determine what constitutes an

effective biomonitor species, including a comprehensive knowledge of the basic biology

and physiology of the species. Establishing baseline hematological, plasma chemistry and

morphological parameters can help ﬁrlﬁll this requirement of an effective biomonitor

Species. With rapid population growth, urbanization and an expanding economy, Uganda

needs effective methods to assess the quality of its environment. Development of suitable

bioindicator species are part of a multifaceted approach to effectively assess

environmental changes and to monitor the impact of such changes on wildlife.

In addition to providing baseline physiological data to create potential

biomonitors, proactive data collection on raptor populations in their natural habitats is

preferable to data collected on small numbers of captive specimens. The latter has

41 become necessary for some raptor species, such as the Spanish Imperial Eagle (Aquila adalberti), due to declines in the wild population (Garcia-Montijano et a1 2002).

Much of the information on African fish eagle behavior, biology and physiology is anecdotal, dated, site specific and non-standardized. There are no known reports describing plasma chemistry or hematological parameters in adult or nestling African fish

eagles. Morphometric data on adult African ﬁsh eagles, to the best of our knowledge, is

not published in the scientiﬁc literature. This paper presents packed cell volumes (PCV),

plasma chemistry values, blood parasite analysis, and morphometric data on nestling and

adult African ﬁsh eagles of known sex. The sampling of African ﬁsh eagles fer the

parameters reported in this paper formed part of a larger study with objectives to

determine concentrations of organochlorinated pesticides and total polychlorinated

biphenyls (PCBS) ﬁ'om eagle plasma, total mercury content in feathers, characterize eagle

nest site habitat and assess the species potential as a biomonitor. The objective of this

paper is to provide physiological data that may prove useﬁrl in the development of this

species as a biomonitor, as well as facilitate the conservation and management of captive

and wild populations of African ﬁsh eagles.

Materials and Methods

African ﬁsh eagles were sampled at Lake Mburo, a six km long freshwater lake in

south western Uganda (0°. 39' S, 30°. 57' E) situated in a 256 km2 national park, and on

Lake Victoria near Entebbe, from Nfo Island (0°. 00' N, 32°. 26' E) to Kisubi Bay (0°. 05'

N, 32°. 35' E). Fish eagles were sampled at Lake Mburo in July, August and December

42 2002 and at Lake Victoria in August 2002 and January 2003. Thirty-three eagles were sampled: ten adults and eight nestlings from Lake Mburo and ﬁve adults and ten nestlings from Lake Victoria. Forty ﬁve percent of birds were sampled between 0600 and 1200 hours and 55% between 1200 to 1800 hours.

Adult ﬁsh eagles were captured on water using a ﬁsh “snare vest” technique.

Tilapia (Oreochromis niloticus) were fitted with fishing line snares (loops) and packed with foam so the fish floated laterally. A total of eight to twelve 5 - 6 cm diameter snares with a free end of line were made per fish. The free ends of line penetrated the body of the fish and were then tied on themselves and the excess line cut. The line was attached to a hand held reel and the fish placed in water. Once captured, the eagle was retrieved and secured by the legs in the tibiotarsal region. Fish eagles swim well so there was little risk of drowning. On Shore, the eagles were placed in dorsal recumbency and the eyes covered. Ten ml of blood were collected fi’om the brachialis vein via a 21 or 23 gauge x

1.9 cm ("/4 inch) butterﬂy catheter (Surﬂo Winged Infusion Set, Elkton, MD) connected to a 10 ml syringe (Luer Lok Tip Syringe, Becton Dickinson and Company, Rutherford,

NJ) ﬂushed with sodium heparin (100 IU/ml). The blood was immediately transferred to a 10 ml lithium heparin vacutainer. An additional 4 ml of blood was drawn and placed in a 5 ml EDTA vacutainer (Becton Dickinson, Franklin Lakes, NJ). Three blood smears were made with fresh blood using the slide on slide technique (Campbell, 1988). Fresh whole blood was also used to determine blood glucose levels (Medisense 2” card glucometer utilizing precision plus sensors”, MediSense Inc, Bedford, MA). A drop of whole blood was placed on a commercially prepared paper sample card for molecular sex

43 determination based on total erythrocyte DNA (Avian Biotech International, Tallahassee,

FL). Five whole breast feathers were hand plucked for determination of total mercury concentrations. A physical examination including scoring body condition (based on pectoral muscle mass and feather condition), whether the crop was empty or full and a visual description of any abnormalities were made. Body measurement methods used were the same as those described for the bald eagle (Bortolotti 1984 a,b). Length of the eighth primary feather and footpad were determined with a 60cm ruler (Figure 3.2)

(Pickett brand Model ASE 24, Forestry Suppliers, Jackson, MI). Hallux and culmen length, as well as bill depth were measured using a dial caliper (Figure 3.3.) (model SPI

2000, Forestry Suppliers, Jackson, MI). Birds were banded with 18-22mm internal diameter metal rivet bands inscribed with a three letter sequential code and the word

“MAKERERE” (Gey Band and Tag Company, Norristown, Pennsylvania, USA). The bands were colored either red or gold for Lake Mburo and black for Entebbe. Suspected female birds were banded on the left leg and suspected males on the right leg. Lastly, birds were placed in a cotton sack and body weight recorded by a spring balance with gradations of 100g (Horns model 20, Douglas Horns Corp, Belmont, CA). Eagles were then released fi'om land at the closest point possible to the capture location. Average time from capture to release was 34 minutes (range 20-45 minutes). African fish eagles were classified as adult if they had attained firll adult plumage color (i.e. were at least five years old).

African ﬁsh eagle nestlings were retrieved for sampling from the nest using professional tree climbing methods (USDA Forest Service, 1996). The main method of tree ascent was using tree climbers (Klein Tools, Chicago, ILL). Eagle nestlings were

gently coaxed to the side of the nest using an “eagle hook” modiﬁed from a car aerial or

ice gaff. Eagle nestlings were placed singly into a ventilated nylon bag and lowered to the

ground for sampling. Sampling of nestlings was as described for adults with the

exception that the volume of blood collected varied from four to 14 ml depending on

body weight. Eagles were aged to within +/- 3 days based on body weight and the

calculations presented by Sumba (1988).

Samples were placed in a chilled cooler. Time of sampling to storage of plasma in

liquid nitrogen was 3.5 hours (range 2-9 hours). Packed cell volume and total plasma

protein (TPP refrac) were determined in the ﬁeld. Microhematocrit capillary tubes (2) were centrifuged at 3000 rpm for 5 minutes (Vulcon Mobilespin PS126-6, Vulcon

Technologies, Grandview, MO) and an average PCV reading recorded. Total plasma

protein was determined using a temperature compensated refractometer (Leica Inc.

Optical Products Division, Buffalo, NY). The remaining blood was centrifuged at 3000

rpm for ten minutes. Plasma was observed visually for hemolysis, icterus, and lipemia

and these changes subjectively classiﬁed as slight, moderate or severe. Plasma was

pipetted into ﬁve 2 m1 cryovials (Cryogenic Vial, Corning Incorporated, Corning, NY)

and deposited into a MVE Doble-20 Vapor Shipper/Liquid Nitrogen Tank (MVE Bio-

Medical Systems, Burnsville, MN). Plasma samples were transported to the Diagnostic

Center for Population and Animal Health (DCPAH) at Michigan State University

Veterinary Medical Center (MSU) then transferred to a — 80°C freezer until analyzed.

Analysis occurred ﬁve months after sampling for 14 of the samples and less than one

45 month for the remainder. Plasma chemistry analyses were performed at the clinical pathology and endocrinology laboratories of the DCPAH at MSU. The plasma chemistry analyses were performed on an Olympus AU640 chemistry analyzer, (Olympus America

Inc., Irving, TX). The electrolyte analyses were performed with a sodium potassium crown ether membrane while the chloride analysis employed a molecular oriented polyvinylchloride membrane. Calcium (Ca), phosphorous (P), TPP col, albumin, aspartate transaminase (AST), creatine kinase (CK), cholesterol and uric acid were performed using olympus reagents. Sodium/potassium ratio, and globulin were calculated from the measured parameters. Total plasma thyroxine (TT4) was measured by two methods: a commercial radioimmunoassay (TT4 RIA) (Diasorin Inc., Stillwater, MN) and a colorometric ELISA (TT4 ELISA) specifically designed to assess thyroid firnction in a range of animal species, including birds (Oxford laboratories, Oxford, MI. Not commercially available at the time of publication). Total plasma triiodothyronine (TT3) was measured using an assay prepared in house at the endocrinology laboratory, DCPAH at MSU. Molecular sex differentiation used a polymerase chain reaction based on the first gene of the avian W chromosome (CHD) (Ellegren, 1996; Griffiths et a1 1998).

Blood smears for blood parasite analysis were stored unstained in slide boxes.

Prior to examination, they were ﬁxed in absolute methanol and then stained with giemsa.

Each slide was examined in its entirety (250X), then for ten nrinutes (SOOX) and for an additional ten minutes under oil immersion in two ﬁve minute sessions (1250X). Both red and white blood cells were examined. The degree of parasitemia was expressed as the percentage of erythrocytes infected per 1000 erythrocytes counted. Parasitological

46 analysis was conducted at the Michigan Department of Natural Resources Rose Lake

Pathology Laboratory.

Analysis of variance was conducted to assess the association between the risk factors of site, age (nestling or adult), gender and plasma chemistry parameters. Analysis of variance was also used to assess the association between morphological characteristics and sex of adult ﬁsh eagles (SAS PROC ANOVA for categorical risk factors, and SAS

PROC GLM for continuous risk factors. SAS 8.2, 2001. SAS Inc., Cary, NC). These analyses were conducted at the univariable (only one risk factor at a time) and multivariable level. Multivariable analyses were conducted to adjust the effect of selected risk factors simultaneously. The level of signiﬁcance (type 1 [a] error) was set at p s 0.05. Descriptive statistics were done using Excel (Microsoft Excel, Microsoft

Corporation, Redmond, WA). An outlying value was deﬁned as being 1.5 times greater or less than the interquartile range. Descriptive statistics are emphasized due to the small sample size. This emulates the methods of other studies examining wild avian hematological and plasma chemistry values where only small sample sizes could be obtained (Garcia- Montijano 2002; Lumsden 1998).

All procedures utilizing birds in this study were carried out under approval of the

Michigan State University All University Committee on Animal Use and Care. The

Uganda National Council for Science and Technology and the Uganda Wildlife Authority granted research permits for this project.

47 Results

Plasma chemistry values and PCVs are presented for adult (Table 3.1.) and nestling

(Table 3.2.) African ﬁsh eagles. Morphometric data are presented for adult African ﬁsh eagles divided by sex (Table 3.3.). Results of analysis of variance of the morphological data are presented in Table 3.4. Results of analysis of variance of the plasma chemistry

data are presented in Table 3.5. Results of multivariable analysis of variance of PCV in

African ﬁsh eagle nestlings is presented in Table 3.6. Two of the plasma samples, a

nestling from Lake Victoria and an adult ﬁ'om Lake Mburo were slightly hemolyzed.

Two plasma samples showed slight (nestlings ﬁ'om Lake Mburo), two moderate

(nestlings from Lake Victoria) and one sample severe lipemia (nestling from Lake

Mburo). Three nestlings from Lake Victoria had an empty crop and seven had full crops.

Four nestlings from Lake Mburo had empty crops and four had full crops. Three adults

from Lake Victoria had full crops and two had empty crops. Six adults from Lake Mburo

had empty crops and four had full crops. All birds were in good body condition as

assessed by pectoral muscle mass and general physical examination. There was a weak

positive correlation between blood glucose levels and a full crop (r2 = 0.012). The mean

PCV for nestlings was 0.27 L/L. There was no signiﬁcant difference (p 2 0.05) in PCV of

nestling ﬁsh eagles of different body weights. Plasma chemistry values that were

signiﬁcantly (p s 0.05) different between adults and nestlings were AST, phosphorous,

potassium and CK, all of which were lower in adults, except AST. Aspartate

transaminase values were signiﬁcantly (p s 0.05) higher in adults than nestlings. There

were signiﬁcant differences (p s 0.05) in cholesterol, albumin, potassium, phosphorous,

TT3 and TT4 RIA values in ﬁsh eagles between the study sites. There was a weak

48 positive correlation (r2 = 0.032) for TT4 levels measured by radioimmunoassay and

ELISA methods. The ratio of TT4 ELISA/T T3 was 4.74.

Total plasma protein values were consistently higher when measured using a temperature compensated refractometer in the ﬁeld than with a colorimetric method in the laboratory.

Mean values for nestlings were 9g/L higher and adult values 10 g/L higher for measurements taken with a refractometer. A strong positive correlation existed between values returned by each method (r = 0.76).

One female nestling (estimated age 52 i 3 days) from Lake Mburo had moderate numbers of erythrocytes (6%) infected with a Plasmodium circumﬂexium like parasite.

Two nestling siblings (estimated ages 42 and 44 i 3 days), a male and female from Lake

Mburo, had light burdens (2% erythrocytes infected) of the same parasite. The nestlings infected with blood parasites also had old, healing digital abrasions, showed the poorest body condition of all the birds examined and had moderate burdens of an unidentiﬁed

species of lice and hippoboscid ﬂies. One infected bird was sampled at 0900, the second at 1200 and the third (with the heaviest burden) at 1300.

Body weights, culmen, footpad, 8‘” primary feather length, and bill depth

parameters were signiﬁcantly (p s 0.05) greater in adult female than in adult male ﬁsh

eagles. Female ﬁsh eagles were on average 20% heavier than male ﬁsh eagles. The ratio of male to female culmen, footpad, 8‘“ primary feather length, and bill depths were 0.37,

0.89, 0.91 and 0.92 respectively.

49 Discussion

Packed cell volumes for nestlings in this study were markedly lower than nestling

values reported in other eagle species (Redig, 1993; Bowerman 2000; Hoeﬂe et al 2000).

The mean PCV of African ﬁsh eagle nestlings in this study was 0.27 L/L. Values

reported for wild bald eagle nestlings are 0.34 L/L (Redig, 1993) and 0.32 L/L

(Bowerman 2000). Values reported for wild Spanish imperial eagles Aquila adalberti were also 0.32 L/L (Hoefle et al., 2000). Despite Afiican fish eagles having a similar

ﬂedging period as these two eagle species, birds of a similar age to those sampled in the

above studies still had lower PCVs. Despite the small nestling sample Size (n =18), there

were no low outlying values (greater than 1.5 times lower than the 25‘” percentile of the

sample) to confound the results. Only three of 18 nestlings had internal or external

parasite burdens that may have contributed to the low PCV. Exclusion of nestlings with

blood parasites from the sample size did not signiﬁcantly raise the mean nestling PCV

(27.44 to 28.13). However, the nestling with the greatest parasite burden had the lowest

recorded PCV of all birds sampled (0.20 LIL). The two nestlings with light parasite

burdens had PCVs equal to the 25‘” percentile of the total nestling sample size. All

nestlings were considered to be well nourished by visual examination and no nestlings

were considered in poor body condition. The ﬁnding that there was no Signiﬁcant

association between PCV and body weight is surprising. Although there may be some

variation due to factors such as nutrition, there should be a positive correlation between

body weight and age over the period of ﬁedging growth (Sumba 1988). Therefore, PCV

should increase with age. Due to the small sample size, results may not be statistically

50 valid as PCV would be expected to increase with age, as reported in studies on captive white storks, Ciconia ciconia (Montesinos et a1 1997) and psittacines (Clubb et a1 1991).

It may be that rapid increases in PCV only occur once nestlings have an increased oxygen demand at the time of fiedging (Hawkey et al 1984). Few studies have examined PCV in relation to nestling age in wild raptors. The reason for the lack of studies may be the greater number of studies conducted on species from temperate regions. Raptors from temperate regions are likely to have well defined breeding seasons, hence nestlings sampled in a particular temporal and spatial period would be expected to be of a comparable age. Such is not the case with Afiican fish eagles from tropical climates where populations may contain eagle pairs with eggs, just hatched young or nestlings about to fledge at the same time.

Sex based variation in AST levels has been recorded in some avian species (Gee et al 1981). The variation in potassium values between adults and nestlings and the

absolute values are similar to those reported for bald eagles (Bowerman 2000; Redig

1993). Red blood cell lysis may precipitate elevated extra-cellular potassium levels

(Fudge 1994). Variation in CK levels is most likely indicative of muscle damage or

injection site trauma during sampling, although overtly, struggling was usually minimal.

Due to the small sample size, the data presented may not be statistically signiﬁcant thus

caution should be exercised with interpretation.

Total plasma thyroxine and triiodothyronine values presented should be

interpreted cautiously. Serum levels of thyroid hormones and the ratio of TT4 to TT3 can

51 vary greatly between genus, species and individuals (Rosskopf et a1 1982). This is

illustrated by the ratio of TT4 ELISA/T T3 being 1.79 for marabou stork nestlings and

4.74 for ﬁsh eagle nestlings in this study. The short half lives of avian TT3 and TT4 can

lead to signiﬁcant diurnal variation in plasma concentrations of these hormones.

Hyperthermia and stress may also reduce serum TT4 concentrations (Rae 2000). Birds

have relatively low concentrations of TT4 compared to humans, so human TT4 RIA may

have low sensitivity in avian species. In addition, commercially available human

radioimmunoassay kits that have not been standardized in a particular laboratory by using

serum from euthyroid birds are difﬁcult to compare between laboratories, different

commercial kits and different species (Merryman and Buckles 1998).

The ﬁnding that TPP values were consistently higher when measured using a

temperature compensated refractometer in the ﬁeld than with a colorimetric method in the laboratory may support studies that demonstrated poor reproducibility of values using

refractometry to determine avian plasma protein values (Lumiej and Maclean 1996). This

may indicate that refractometric methods should be used only as an approximation of

plasma protein levels.

Bennett et a1 (1977) recorded blood parasite prevalence in a variety of avian

species (not including African ﬁsh eagles), over 5 years at Zika Forest (about 30 km from

Kampala). They found an overall prevalence of 34.8 % with the vast majority of

infections being Haemoproteus spp. Plasmodium spp. infection comprised only

approximately two percent of all infections and there appeared to be a slight positive

52 correlation between parasite prevalence and rainfall. Our samples were taken during the dry seasons.

Little morphometric data exists on African ﬁsh eagles. Average body weights in this study are similar to those cited by Sumba (1988) for captive ﬁsh eagles at Entebbe of

2.250kg for males (n = 3) and 2.835 kg (n = 2) for females. Body weight ranges presented by Brown (1980), 3.000-3.600kg (n = 3) for females and 1986-2.497 kg (n =

4) for males, fall within ranges found in this study. The morphometric data supports the conclusion that, like other Haliaeetus species, females have a larger body mass than males.

It is hoped that the information presented in this paper can be utilized as a foundation on which further research can construct a comprehensive biological and physiological database for the African ﬁsh eagle. Such a database may prove a valuable diagnostic tool for the conservation of individual species, and the management and monitoring of ecosystem health, through use of the species as a biomonitor. Such long- terrn conservation goals are necessary and warranted on the African continent.

53 References

Bennett GF, White ME & Williams NA. 1977. Additional observations on the blood parasites of Ugandan birds. J. Wildl. Dis. 13: 251 — 257.

Bowerman WW, Best DA, Giesy JP, .Meyer MW, Shieldcastle JP, Postupalsky S & Sikarskie JG. 2003. Associations between regional differences in polychlorinated biphenyls and dichlorodiphenyldichloroethylene in blood of nestling bald eagles and reproductive productivity. Environ. Toxicol. Chem. 22: 371-376.

Bowerman WW, Stickle JE, Sikarskie JG & Giesy JP. 2000. Hematology and serum chemistries of nestling bald eagles (Haliaeetus leucocephalus) in the lower peninsula of Michigan, USA. Chemosphere 41: 1575-1579.

Brown L. 1980. The African Fish Eagle. Bailey Bros. & Swinfen Ltd., Folkestone, England. Pp. 7-11.

Bortolotti GR 1984a. Criteria for determining age and sex of nestling bald eagles. J. Field Omith. 55: 467-481.

Bortolotti GR. 1984b. Sexual dimorphism and age related size variation in bald eagles. J. Wildl. Manage. 48: 72-81.

Campbell, T.W. 1988. Avian Hematology and Cytology. 1988. Iowa State University Press. Ames. Iowa. pp. 3-17.

Clubb SL, Schubot RM, Zinkl JG, Wolf S, Escobar J & Kabbur MB. 1991. Hematologic and serum biochemical reference intervals in juvenile cockatoos. J. Assoc. Avian Vet. 5:16-26.

Ellegren H, 1996. First gene on the avian W chromosome (CHD) provides a tag for universal sexing of non-ratite birds. In: Proc. Roy. Soc. Lon. Series B, Biol. Sc. 263: 1635-1641.

Fudge AH. 1994 Blood testing artifacts: Intrepretation and prevention. Semin. Avian Exotic Pet Med. 3: 2-4.

54 Garcia-Montijano M A, Lemus JA, Montesinos A, Canales R Luaces I & Pereira P. 2002. Blood chemistry, protein electrophoresis, and hematological values of captive Spanish imperial eagles (Aquila adalberti). J. Zoo & Wild. Med. 33: 112-117.

Gee GF, Carpenter JW & Hensler L. 1981. Species differences in hematological values of captive cranes, raptors and quail. J. Wildl. Manag. 45: 463-483.

Griffiths R, Double MC, Orr K & Dawson RJ. 1998. A DNA test to sex most birds. Mol. Ecol. 7: 1071-1075.

Hawkey, CM, Hart MG & Samour J. 1984. Age related hematological changes and hematological responses in Chilean ﬂamingoes (Phoeconipterus chiliensis). Avian Path. 13 2223-228.

Hoeﬂe U, Blanco JM, Sauer-Guerth H & Wink M. Molecular sex determination in Spanish imperial eagles (Aquila adalberti) nestlings and sex related variation in morphometric parameters. In: Raptor Biomedicine III. 2000. JT Lumeij, JD Remple, PT Redig, M Lierz &; JE. Cooper ((Eds.).) Zoological Education Network, Lake Worth, Florida. pp 289-293.

Lumeij J & Maclean B. 1996. Total protein determination in pigeon plasma and serum: comparison of refractometric methods with biuret method. J. Avian Med. & Surg. 10:150-152.

Lumsden JH. 1998. “Normal” or reference values: questions and comments. Vet. Clin. Pathol. 27: 102-106.

Merryman JI & Buckles EL. 1998. The Avian thyroid gland. Part two: a review of function and pathophysiology. J. Avian Medic. Surg. 12:238-242.

Montesinos AA, Sainz MV, Pablos F, Mazzuchelli & Tesouro MA. 1997. Hematological and plasma biochemical reference intervals in young white storks. J. Wildl. Dis. 33: 405-412.

55 Rae M. 2000. Avian endocrine disorders. In: Laboratory medicine avian and exotic pets. AM.Fudge (ed.). W.B.Saunders and Company Ltd. Philadelphia. Pennsylvania. Pp. 78- 82.

Redig PT. 1993. Medical Management of Birds of Prey. 3“’ Ed. The Raptor Center at the University of Minnesota, St Paul, Minnesota. Pp. 35-44.

Rosskopf WJ, Woerpol RW & Rosskopf G 1982. Normal thyroid values for common pet birds. Vet. Medic. /Small Anim. Clinic. 77: 409-412.

Stewart K M, Matthiesen DP, Leblanc L & West J. 1997. Prey diversity and selectivity of the Afiican fish eagle: data fiom a roost in northern Kenya. Afr. J. Ecol. 35:133-145.

Sumba SJA. 1988. Nestling grth in the African ﬁsh eagle in Uganda. Afr. J. Ecol. 26: 315-321.

United States Department of Agriculture Forestry Service. National Tree Climbing Field Guide, USDA Forest Service 1996, Missoula Technology and Development Center, Missoula, Montana, 1996.

56 Table 3.1: Packed cell volumes and selected plasma chemistry values and morphometric measurements, from adult African ﬁsh eagles (Haliaeetus vocifer) sampled at Lake

Mburo and Lake Victoria near Entebbe, Uganda (n =15).

’ PCV = packed cell volume b TPP reﬁ'ac = total plasma protein measured by refractometry

° TPP Col = total plasma protein measured by a colorimetric method

" TT4 radioimmunoassay = thyroxine radioimmunoassay

°TT4 elisa = thyroxine enzyme linked immunoabsorbent assay

{TT3 = triiodothyronine

g AST = aspartate transaminase

h CK = creatine phosphokinase iP = phosphorous jNet/K = sodium/potassium ratio

" = determined on whole blood

57 Table 3.1 Packed cell volume, plasma chemistry values and morphometric measurements from adult African ﬁsh eagles

Drean Median-SD or . Q3 Min Max

' PCV UL 0.45 0.45 0.0225 0.44 0.46 0.43 0.53 ”TPP Refrac g/L 48 47 4.7 45 52 38 54 ‘TPP Col g/L 38 38 4 35 41 30 42

‘TT4 nMol/l 4.4 4.0 2.03 3.0 5.5 1.0 8.0 e'1'r4 Elisa nMol/l 14.56 12.50 8.61 9.25 16.20 6.60 41.60 'm nMol/l 1.6 1.5 0.41 1.2 1.7 1 2.4

“Glucose mMol/L 12.4 11.9 2.01 10.7 12.8 10.3 16.3 'AST U/L 194 152 117 139 199 121 590 Calcium mMol/L 2.4 2.4 0.13 2.34 2.50 2.12 2.57 "CK U/L 217 215 53 184 252 127 320 ’P mMol/L 0.74 0.68 0.30 0.48 0.92 0.42 1.45

Uric Acid mMol/L 0.998 1.011 0.408 0.690 1.190 0.291 1.731 Albuming/L 12 12 1.3 12 13 11 15 Globulin g/L 25 26 3.3 23 28 19 30 Sodium mMol/L 153 155 5.56 153 155 143 161 Potassium mMol/L 1.3 1.2 0.33 1.1 1.5 1.0 1.8

’Na/K ratio 123 127 31.67 95 144 87 161

Chloride mMol/L 115 116 5.21 112 117 105 124 Cholesterol mMol/L 4.69 4.56 0.644 4.22 5.05 3.86 6.16

58 Table 3.2: Packed cell volumes and selected plasma chemistry values from nestling

African ﬁsh eagles (Haliaeetus vocifer) sampled at Lake Mburo and Lake Victoria near

Entebbe, Uganda (n =18).

" PCV = packed cell volume b TPP refrac = total plasma protein measured by refractometry

° TPP Col = total plasma protein measured by a colorimetric method

" TT4 RIA = thyroxine radioimmunoassay

° TT4 ELISA = thyroxine enzyme linked immonoabsorbent assay fTT3 = triiodothyronine g AST = aspartate transaminase h CK = creatine kinase iNa/K = sodium/potassium ratio

’P = phosphorous k = determined on whole blood

59 Table 3.2. PCV and plasma chemistry values from nestling African ﬁsh eagles

Mean Median SD Q1 Q3 Min Max

Age days 27 22 12 18 37 9 52 Weight Kg 1.45 1.33 0.63 0.91 1.94 0.50 2.60 ‘PCVL/L 0.27 0.28 0.03 0.26 0.30 0.20 0.33

”TPP Refracg/L 45 45 4 44 48 36 52 ‘TI'PColg/L 36 37 4 32 39 27 42

‘TT4RIAnMol/l 8 7 4 5 11 2 15 ‘TT4elisanMol/l 11.44 10.50 2.48 9.80 13.20 8.10 16.00 'r'rs nMon 2.28 2.30 0.43 2.10 2.60 1.40 2.80 “Glucose mMol/L 13.8 14.3 1.8 12.6 14.9 10.4 17.7 'AST U/L 123 120 34 95 143 75 185 Calcium mMol/L 2.62 2.65 0.1 2.60 2.70 2.40 2.80 I'CKU/L 906 754 515 517 1201 178 1880

’PmMol/L 1.97 1.68 0.84 1.55 2.13 0.77 3.91 Uric Acid rnMol/L 0.898 0.922 0.345 0.660 1.041 0.422 1.65 Albuming/L 13.8 13.5 2.1 12.3 16.0 10.0 17.0 Globuling/L 22.0 21.0 3.6 20.3 24.5 16.0 28.0 Sodium mMol/L 148 148 2 147 149 145 154 Potassium mmol/L 2.45 2.30 0.65 2.10 2.90 1.30 3.80

’Na/Kratio 65 65 19 51 71 38 112 Chloride mMol/L 110 111 5 107 113 101 118

Cholesterol mMol/L 5.49 5.36 1.11 4.76 6.24 3.29 7.41

60 Table 3.3: Morphometric measurements of adult male and female Aﬁican ﬁsh eagles (Haliaeetus vocifer) (n = 15) from Lake Mburo and Lake Victoria near Entebbe.

61 Table 3.3. Morphometric measurements of adult male and female African ﬁsh eagles

Adult Male Fish Eagles (n=9)

Mean Median SD Q1 Q3 Min Max

Weight (Kg) 2.3 2.3 0.1 2.3 2.4 2.1 2.5

8'll Primary (mm) 380 387 16 380 389 340 391

Footpad (mm) 109 110 5 106 112 101 116

Bill Depth (mm) 25.24 24.76 1.13 24.64 24.88 24.49 27.81

Culmen (mm) 39.97 39.54 0.94 39.50 40.85 38.35 41.28

Hallux (mm) 35.64 36.17 1.87 35.15 36.40 31.29 38.02

Adult Female Fish Eagles (n=6)

Weight(Kg) 3.0 2.9 0.4 2.7 3.4 2.5 3.6

8" Primary (mm) 417 420 8 412 424 405 425

Footpad(mm) 123 124 5 120 125 115 128 Bill Depth (mm) 27.39 27.21 1.79 26.39 28.29 25.05 30.11

Culmen(mm) 45.71 44.31 5.14 43.35 44.96 41.55 55.89 Hallux(mm) 39.90 39.76 1.89 38.50 40.96 37.70 42.71

62 Table 3.4. Univariable analysis of variance of morphological data by gender from African ﬁsh eagles from Lake Mburo and Lake Victoria near Entebbe (n=15)

Variable Gender Mean (sd) F

' ’ "_ “W" m F818;; F 6 3.02145) Body weight (kg) 18.34 .0009 Male 2.36 (.12)

m . Female 417.50 (8.22) 8 prrmary feather length (mm) 26.91 .0002 380.56 (15.95) Male \OO

122.67 (4.76) Female 0‘ Footpad length (mm) 28.30 .0001 Male 109.22 (4.82)

Female 27.39 (1.79) Bill depth (mm) 8.30 .0129 Male 25.24 (1.13)

Female 45.71 (5.14) Culmen length (mm) 11.07 .0055 Male 39.97 (.94)

Female 39.90 (1.89) Hallux length (mm) 18.54 .0009 Male 35.64 (1.87)

63 Table 3.5: Analysis of variance of plasma chemistry values in adult (r1 = 15) and nestling

( n = 18) African ﬁsh eagles (Haliaeetus vocifer) from Lake Mburo and Lake Victoria near Entebbe, Uganda.

" AST = aspartate transaminase b C hol. = cholesterol

° CK = creatine kinase dPCV = packed cell volume eP = phosphorous fTT4 ELISA = thyroxine enzyme linked immunoabsorbent assay g TT4 RIA = thyroxine radioimmunoassay

" rr3 = triiodothyronine

’ TPP (ref) = total plasma protein determined by refractometry j TPP (col) = total plasma protein detemiined by a colorimetric method

" = determined on whole blood

64 Table 3.5. Analysis of variance of plasma chemistry values in ﬁsh eagles (n = 33)

Plasma Variables F p Plasma Variable F p, I. H I £99914!!! A A_ A A _ A A Chemistry

Site .82 .3723 Site 17.17 .0003 Age 5.84 .0222 Age 10.55 .0029 “AST Gender .12 .7293 A'b‘m'i“ Gender .66 .4222

Overall 2.26 .1024 Overall 9.46 .0002

Site .74 .3980 Site 9.03 .0054 Age 27.45 < .0001 Age 35.60 < .0001 °cr< Gender 2.45 .1282 cam” Gender 0.0 .9724

Overall 10.21 < .0001 Overall 14.88 < .0001

Site 9.84 .0039 Site .38 .5405 Age 7.82 .0091 Age 7.22 .0118 how" Gender .05 .8270 Ch'm'id‘ Gender .32 .5753 Overall 5.90 .0028 Overall 2.64 .0682

Site 3.44 .0739 Site .73 .4008

Age 7.42 .0108 Age 4.62 .0402 G'°b““" Gender .74 .3979 “alum” Gender .20 .6591 Overall 3.87 .0193 Overall 1.85 .1608

65 Table 3.5 (continued)

Plasma Variables F Plasma Variables F Chemistry Chemistry

Site 6.51 .0163 Site .32 .5744

Age 50.21 < .0001 Age 7.61 .0099 Potassium Sodium Gender 0.0 .9536 Gender .55 .4628

Overall 18.91 < .0001 Overall 2.83 .0559

Site 10.40 .003 l She .25 .6225

Age 39.30 < .0001 Age 275.30 < .0001 “PCV Gender .79 .3824 Gender 1 .59 .2181

Overall 16.83 < .0001 Overall 92.38 < .0001

Site 17.76 .0002 Site 20.20 .0001

Age 35.56 < .0001 Age 16.55 .0003 ‘TT4 RIA Gender .72 .4043 Gender .36 .5523

Overall 18.01 < .0001 Overall 12.37 < .0001

Site .92 .3451 Site 3.71 .0638

2.07 .1608 Age .65 .4250 f'r'r4 ELISA Uric Acid Gender 1.40 .2474 Gender 3.73 .0633

Overall 1.46 .2456 Overall 2.70 .0641

Site 2.36 .1352 Site .01 .9347

Age 2.99 .0943 Age 1 .54 .2249 lTPP (Ref) Gender .03 .8631 (Col) Gender .62 .4381

Overall 1.79 .1703 Overall .72 .5477

66 Table 3.6. Univariable and multivariable analysis of variance of packed cell volume in African ﬁsh eagle (n=18) nestlings.

Species Variable Level 11 Mean (sd) F p

Lake Victoria I0 2981-1199;“ I Site 22.29 .0002 Lake Mburo 8 24.5 (2.78)

Bodyweight - 18 - 2.35 .1445

Fish Female 7 26.71 (4.57) eagle Gender .47 .5033 Male 11 27.91 (2.88) nestlings

Site - - 21.39 .0004

Multivariable Body Weight - - .18 .6759

ANOVA Gender - - 1.17 .2982

Overall - - 7.58 .0030

67 Figure 3.1: Technique for measuring Aﬁican ﬁsh eagle footpad length with a rigid ruler.

Tendons must be stretched to obtain maximum extension of footpad for accurate reading.

(Modiﬁed from Bortolotti 1984 a, b).

Figure 3.2. Technique for measuring African ﬁsh eagle culmen length (A), bill depth (C) and hallux length (D). (Modiﬁed from Bortolotti 1984 a, b).

68 Figure 3.1. Technique for measuring African Fish Eagle footpad length (Modiﬁed from Bortolotti 1984 a, b) l

ﬂ 4

I ’ : F 1 x 1. I w l' 9" t ‘. 'r' . s -.' ' "o .‘ i .s . .. ’c . in _9 A. l I ‘-- C ‘ .

Figure 3.2. Technique for measuring African ﬁsh eagle culmen length (A), bill depth

Chapter 4

Packed Cell Volume, Plasma Chemistry Values and Survey for Blood

Parasites in Nestling Marabou Storks (Leptoptilos crumemjferus) in

Uganda

70 Abstract

The purpose of this research was to establish baseline physiological parameters, such as plasma chemistry values in marabou storks and assess the feasibility of developing this species as a biomonitor of environmental change. Packed cell volumes

(PCV) and plasma chemistry parameters were measured in twenty nestling marabou storks (Leptoptilos crumeniferus) in January 2003. Marabou nestlings were part of a colony located in the center of Kampala, Uganda. Nestlings were also surveyed for the presence of blood parasites. Sex was genetically determined using erythrocyte DNA. No blood parasites were found. There were no signiﬁcant differences (p 2 0.05) in plasma chemistry values or PCV between sexes. Total plasma protein (TPP), uric acid, phosphorous and creatine kinase were generally higher relative to published data on other avian species, including nestling white storks (Ciconia cicom'a). Thyroxine levels measured by both ELISA and radioimmunoabsorbent assay were variable between nestlings. All other values were similar to those reported for a variety of avian species.

Introduction

The marabou stork is indigenous to tropical Africa where it is common to abundant in most parts of its range. The marabou stork (Leptoptilos crumeniferus) has responded to increasing urbanization and centralization of human populations by adopting a scavenging lifestyle and cosmopolitan diet in urban areas. Populations have increased in Kampala and breeding colonies may be found in the city center (Hancock et a1 1992). This research formed part of a study examining persistent organic pollutant concentrations in marabou storks and Aﬁican ﬁsh eagles (Haliaeetus vocifer) and

71 assessing their potential as biomonitor species. To be an effective biomonitor, information is required on the biology and physiology of a species. The biology of the marabou stork in Uganda has been well described by Pomeroy (1977, 1978). Adult marabous weigh between 5 - 8 kg and the diet of the birds in Kampala probably includes almost anything organic, such as garbage, ﬁsh remains, abattoir refuse and a large amount of vegetable matter (Brown et al 1982). However, during the period of nestling grth increased amounts of protein are taken in the form of ﬁsh, frogs and rodents.

Marabous undertake short, mainly north south migrations in Uganda, coinciding with rainfall seasonality (Pomeroy, 1978 a, 1978 b). Marabous are colonial nesters with multiple nests being built in particular tree types, such as Mvule (Chlorophora excelsa) and Tabebuia pentaphylla. The incubation period is 30.3 days and the average clutch size is 2 - 3 eggs. Marabous have an extremely long period from hatching to fiedging, being about 135 days with first flights out of the nest at 110 - 115 days Marabous first breed at

6 - 7 years and can live up to 25 years in captivity. Breeding success is low but has reportedly increased in recent years in Kampala (Hancock et a1 1992). Despite its ubiquity, no plasma chemistry or hematological parameters have been reported for wild marabou storks. The objective of this study was to determine baseline plasma chemistry and packed cell volume (PCV) values for nestling marabou storks.

Materials and Methods

Packed cell volumes, plasma chemistry parameters and body weights were recorded in twenty nestling marabou storks in January 2003. Eleven nestlings were male and nine were female as determined by molecular sexing based on total erythrocyte DNA

72 (Avian Biotech International, Tallahassee, FL). All nestlings were also surveyed for the presence of blood parasites. Marabou nestlings were part of a colony located along Nile

Avenue, in central Kampala, Uganda (0° 19' N 35° 25' E). Kampala is the capital and largest city in Uganda with an estimated population of 1, 208, 544 in 2002 (Uganda

Bureau of Statistics). Marabou stork nestlings were sampled from twelve nests, in six trees (Tabebuia pentaphylla), in one colony, on one day from 0700 till 1800. Marabou

Stork nestlings were temporarily captured for sample collection using professional tree climbing techniques (USDA Forest Service 1996), with modiﬁcations for tropical tree species and environmental conditions. The main method of tree ascent was using tree climbers or occasionally free climbing. When necessary, marabou stork nestlings were gently coaxed to the side of the nest using an ice gaff (Mason Tackle Company, Otisville,

MI) with the point of the hook covered with protective plastic. Parents never left the nest

(and vigorously defended the chicks) or remained close to the nest and returned upon retreat of the climber. Marabou stork nestlings were placed singly into a 40 cm diameter nylon bag (The Taku Tailor, Juneau, AK). The bag had a padded bottom, velcro tab fasteners and multiple ventilation holes. The bag was attached to a rope and lowered to the ground for sampling. Twelve n11 of blood were collected fi'om the medial metatarsal vein via a 23 gauge x 1.9 cm (3/4 inch) butterfly catheter (Surflo Winged Infusion Set,

Terumo, Elkton, MD) connected to a 3 ml syringe ﬂushed with sodium heparin (100

IU/ml) to prevent coagulation. Eight ml of blood were immediately transferred to a 10 ml vacutainer tube containing lithum heparin anticoagulant and the remainder transferred to a 5 ml vacutainer tube containing calcium EDTA anticoagulant (Becton Dickinson,

Franklin Lakes, NJ). Three blood smears were made ﬁom fresh blood using the slide on

73 slide technique (Campbell 1988). Fresh whole blood was also used to determine blood glucose levels with a hand held glucometer (Medisense 23) card glucometer utilizing precision plus sensors“, Medisense Inc. Bedford, MA). A drop of whole blood was placed on a commercially prepared paper sample card for molecular sex determination based on total erythrocyte DNA (Avian Biotech International, Tallahassee, FL). Five whole breast feathers were then hand plucked for determination of total mercury concentrations. A physical examination including scoring body condition (based on pectoral muscle mass and feather condition), whether the crop was empty or full and a visual description of any abnormalities was made. Bill depth, culmen length and bill circumference were measured. Lastly, birds were placed in a cotton sack and body weight recorded by a spring balance with gradations of 100g (Horns model 20, Douglas Horns

Corp. Belmont, CA). Average time for removal ﬁ'om the nest until return was 14.5 minutes (range 7-22).

Samples were placed in a chilled cooler after collection. Separation of plasma occurred on the same day as sampling. Packed cell volume (PCV) and total plasma protein (TPP refrac) were determined in the ﬁeld ﬁ'om blood in the EDTA vacutainer.

Microhematocrit capillary tubes (2) were centrifuged at 3000 rpm for 5 nrinutes (Vulcon

Mobilespin PS126-6, Vulcon Technologies, Grandview, MO) and an average PCV reading recorded. Total plasma protein was determined using a temperature compensated refractometer (Leica Inc. Optical Products Division, Buffalo, NY). The remaining blood in the EDTA and lithium heparin vacutainers was centrifuged at 3000 rpm for ten minutes. Plasma was observed visually for hemolysis, icterus, and lipemia and

74 subjectively classiﬁed as slight, moderate or severe. Plasma was removed and divided into ﬁve 2 ml cryovials (Cryogenic Vial, Corning Incorporated, Corning, NY) that were deposited into a MVE Doble-20 Vapor Shipper/Liquid Nitrogen Tank (MVE Bio-

Medical Systems, Bumsville, MN). Plasma samples were transported to the Diagnostic

Center for Population and Animal Health (DCPAH) at the Michigan State University

(MSU) in the vapor shipper then transferred to a — 80 °C ultra-low freezer until analyzed.

Analysis was done two weeks after sampling. Analyses of plasma chemistry values were performed at the clinical pathology laboratory of the DCPAH at the MSU. The plasma chemistry analyses were performed on an Olympus AU640 chemistry analyzer,

(Olympus America, Inc. Irving TX). The electrolyte analyses were performed with a sodium potassium crown ether membrane while the chloride analysis employed a molecular oriented Polyvinylchloride membrane. Calcium, phosphorous, TPP col, albumin, aspartate amino-transaminase (AST), creatine kinase (CK), cholesterol and uric acid were performed using Olympus reagents. Sodium/potassium ratio, and globulins were calculated from the measured parameters.

Molecular sex differentiation was done using a polymerase chain reaction based on the ﬁrst gene of the avian W chromosome (CHD) (Grifﬁths et al 1998).

Total plasma thyroxine (TT4) was measured by two methods: a commercial radioimmunoassay (TT4 RIA) (Diasorin Inc., Stillwater, MN) and a colorometric ELISA

(TT4 ELISA) speciﬁcally designed to assess thyroid function in a range of animal species, including birds (Oxford laboratories, Oxford, MI. Not commercially available at

75 the time of publication). Triiodothyronine (TT3) was measured using a radioimmunoassay prepared in house at the endocrinology laboratory, DCPAH at MSU.

Blood smears for blood parasite analysis were stored unstained in plastic slide boxes. Prior to examination they were ﬁxed in absolute methanol and then stained with giemsa. Each slide was examined in its entirety (250X), then for ten minutes (500X) and for an additional ten minutes under oil immersion (1250) in two ﬁve minute sessions.

Both red and white blood cells were examined. The degree of parasitemia was expressed as the percentage of erythrocytes infected per 1000 erythrocytes counted. Parasitological analysis was conducted at the Michigan Department of Natural Resources Rose Lake

Pathology Laboratory.

Analysis of variance (ANOVA) was conducted to assess the association between gender, body weight and plasma chemistry parameters of marabou stork nestlings (SAS

PROC AN OVA for categorical risk factors, and SAS PROC GLM for continuous risk factors. SAS Inc., Cary, NC). These analyses were conducted both at the univariable

(only one risk factor at a time) and multivariable level. Multivariable analyses were conducted to adjust the effect of selected risk factors simultaneously. The level of signiﬁcance for type 1 (a) error was set at a probability of 0.05. Descriptive statistics were done using Excel (Microsoft Excel, Microsoft Corporation, Redmond, WA). An outlying value was deﬁned as being 1.5 times greater or less than the interquartile range.

Outlying values occurred for four parameters and were removed ﬁom the results (Table

4.1.). This emulates the methodology of other studies examining wild avian

76 hematological and plasma chemistry values where only small sample sizes could be obtained (Garcia- Montijano 2002; Lumsden 1998).

All procedures utilizing birds in this study were carried out under approval of the

Michigan State University All University Committee on Animal Use and Care. The

Uganda National Council for Science and Technology and the Uganda Wildlife Authority granted research permits for this project.

Results

All birds were in good body condition as assessed by pectoral muscle mass and no abnormalities were detected. The mean weight of the nestlings was 4.1 +/- 1.6 kg. The minimum and maximum nestling weights were 1.65 kg and 7.60 kg. Six birds had full and fourteen had empty crops. Eight of the samples showed slight, two moderate and

ﬁve severe hemolysis.

Results of PCV, plasma chemistry parameters and body weight are presented in

Table 4.1. Analysis of variance for plasma chemistry parameters in nestling marabou storks is presented in Table 4.2. No blood parasites were observed. There were no signiﬁcant differences in plasma chemistry values or PCV between sexes except for globulin and TPP levels. Globulin levels also varied with bird body weight (p s 0.05).

Female nestlings had signiﬁcantly higher (p s. 0.05) TTP levels as measured by refractometry and colorimetric methods. Weight of the nestling was not signiﬁcant when assessed simultaneously with gender for its association with TPP. A strong positive

77 correlation existed between TPP values returned by using a temperature compensated refractometer in the ﬁeld (r2 = 0. 78) and the colorimetric method in the laboratory. Mean values for nestlings were 5 g/L higher when measured by refractometry. There was a positive correlation (r2 = 0.576) between TT4 levels measured by radioimmunoassay to

TT4 levels measured by the ELISA method. The ratio of the sum TT4 ELISA/T T3 was

1.79.

Discussion

Most plasma chemistry parameters and PCVS reported in this study for marabou stork nestlings were similar to those recorded in wild and caged bird nestlings

(Bowerman et al 2000; Redig 1993; Clubb et a1 1993; Hoeﬂe et al 2000). However, TPP, phosphorous, potassium and CK were higher and AST values lower compared to nestlings of other species recorded in the above studies. These same increases were apparent when compared to nestling white stork (C iconia ciconia) plasma chemistry parameters (Montesinos et al 1997), with the exception of phosphorous and potassium, which were similar to values recorded for white storks. White storks were the only other

Ciconiiforrnes in which plasma chemistry values were recorded from nestlings.

Elevations in potassium and phosphorous values were most likely artifactual and caused by hemolysis. Elevated CK values were most likely a combination of sample hemolysis and struggling during handling. A possible explanation for elevated TPP levels in nestlings is that whole vertebrate prey, high in protein is favored by breeding adults during periods of nestling growth (Hancock et al 1992). The parents therefore provide an increased source of dietary protein to the nestlings and this may elevate plasma albumin

78 levels. Another explanation, especially in the colony sampled, is that an urban scavenging lifestyle may supply increased dietary protein for the entire year, irrespective of breeding status. Plasma chemistry samples from breeding, non-breeding and rural colonies would be necessary to validate these explanations.

The variation in TPP values measured by refractometry and the colorimetric method support studies that demonstrate poor reproducibility of values using refractometry to determine avian plasma protein (Lumiej & Maclean 1996). The correlation between both methods of measurement suggest refractometric measurements of TPP provide readings that are consistent but not precise. This may indicate that refractometry should be used only as an approximation, rather than an absolute indicator of TPP.

Differences in AST and Ca levels between sexes have been recorded in avian species (Gee et al 1981), but this has not been a consistent ﬁnding in all species studied and was not found in this study. Apart ﬁom a small sample size, no conclusions can be drawn from the higher globulin and TPP values returned for females

Total plasma TT4 and TT3 values presented should be interpreted cautiously.

Serum levels of thyroid hormones and the ratio of TT4 to TT3 can vary greatly between genus, species and individuals (Rosskopf et al 1982). This is supported by the ratio of

TT4 ELISA/TT 3 for marabou stork nestlings found in this study. The short half- lives of avian TT3 and TT4 can lead to signiﬁcant diurnal variation in measurable values.

79 Hypertherrnia and stress may also reduce serum TT4 concentrations (Rae 2000). Birds have relatively low concentrations of TT4 compared with humans so human TT4 RIA may have low sensitivity in avian species. In addition, commercially available human

RIA kits that have not been standardized in a particular laboratory by using serum from euthyroid birds are difﬁcult to compare between laboratories, different commercial kits and different species (Merryman & Buckles 1998).

The lack of hematozoan parasites in blood smears from nestling marabou storks may be an artifact of a small sample size. The short temporal period over which infections could occur in the nestlings may also be a reason for the negative ﬁndings.

Bennett et al (1977) recorded blood parasite prevalence in a variety of avian species (not including marabou storks), over 5 years at Zika Forest (about 30 km ﬁ'om Kampala).

They found an overall prevalence of 34.8 % with the vast majority of infections being

Haemoproteus spp. There appeared to be a slight positive correlation between parasite prevalence and rainfall. Our samples were taken at the conclusion of a prolonged wet season.

There is a paucity of physiological data on many common African avian species within their natural habitat. Species such as the marabou stork could possibly serve as useﬁil monitors of ecosystem health, in both urban and rural environments. Plasma chemistry parameters will be very useful to help achieve this goal, and to establish a physiological database for the species.

80 References

Bennett GF, White ME & Williams NA. 1977. Additional observations on the blood parasites of Ugandan birds. J. Wildl. Dis. 13: 251 — 257.

Brown L, Urban EK & Newman K. 1982. The Birds of Africa. I. Academic Press, New York.

Campbell TW. 1988. Avian Hematology and Cytology. Iowa State University Press. Ames. Iowa. pp. 3-17.

Clubb SL, Schubot RM, Zinkl JG, Wolf S, Escobar J & Kabbur MB. 1991. Hematologic and serum biochemical reference intervals in juvenile cockatoos. J. Assoc. Avian Vet. 5:16-26. .

Garcia-Montijano M A, Lemus J A, Montesinos A, Canales 1L Luaces I & Pereira P. 2002. Blood chemistry, protein electrophoresis, and hematological values of captive Spanish imperial eagles (Aquila adalberti). J. Zoo & Wild. Med. 33: 112-117.

Gee GF, Carpenter JW & Hensler L. 1981. Species differences in hematological values of captive cranes, raptors and quail. J. Wild. Manag. 45: 463-483.

Grifﬁths R, Double MC, Orr K & Dawson RJ. 1998. A DNA test to sex most birds. Mol. Ecol. 7: 1071-1075.

Hancock JA, Kushlan JA & Kahl MP. 1992. Storks, Ibises and Spoonbills of the World. Academic Press Lim. San Diego, California, USA. Pp. 133-138.

Hoeﬂe U, Blanco JM, Sauer-Guerth H & Wink M. Molecular sex determination in Spanish imperial eagles (Aquila adalberri) nestlings and sex related variation in morphometric parameters. In: Raptor Biomedicine III. 2000. JT Lumeij, JD Remple, PT Redig, M Lierz & JE. Cooper (Eds) Zoological Education Network, Lake Worth, Florida. pp 289-293.

81 Lumeij J & Maclean B. 1996. Total protein determination in pigeon plasma and serum: comparison of refractometric methods with biuret method. J. Avian Med. Surg. 10:150- 152.

Lumsden JH. 1998. “Normal” or reference values: questions and comments. Vet. Clin. Pathol. 27: 102-106.

Merryman JI & Buckles EL. 1998. The Avian thyroid gland. Part two: a review of function and pathophysiology. J. Avian Medic. Surg. 122238-242.

Montesinos AA, Sainz MV, Pablos F, Mazzuchelli & Tesouro MA. 1997. Hematological and plasma biochemical reference intervals in young white storks. J. Wildl. Dis. 33: 405-412.

Pomeroy DE. 1977. The biology of marabou storks in Uganda. 1. Some characteristics of the species, and the population structure. Ardea, 65: 1-24.

Pomeroy DE. 1978. The biology of marabou storks in Uganda. II. Breeding biology and general review. Ardea. 66: 1-23.

Rae M. 2000. Avian endocrine disorders. In: Laboratory medicine avian and exotic pets. AM.Fudge (ed.). W.B.Saunders and Company Ltd. Philadelphia. Pennsylvania. Pp. 78- 82.

Redig PT. 1993. Medical Management of Birds of Prey. 3" Ed. The Raptor Center at the University of Minnesota, St Paul, Minnesota Pp. 35-44.

Rosskopf WJ, Woerpol RW & Rosskopf G 1982. Normal thyroid values for common pet birds. Vet. Medic. /Small Anim. Clinic. 77: 409-412.

Uganda Bureau of Statistics. 2002. Key Economic Indicators. 47th Issue. Ch 39. Uganda Bureau of Statistics, Entebbe, Uganda.

82 United States Department of Agriculture Forestry Service. 1996. National tree climbing ﬁeld guide, USDA Forest Service. Missoula Technology and Development Center, Missoula, Montana, 1996.

83 Table 4.1: Body weight, packed cell volume and plasma chemistry values from nestling marabou storks (Lepiopiilos crumemferus) in Kampala, Uganda (n =20).

8 Q1 = the 25’h percentile of the sample Size b Q3 = the 75"’ percentile of the sample size

° indicates outlying values (greater than 1.5 x the inter-quartile range) have been removed and the sample size adjusted accordingly. dPCV = packed cell volume

° TPP Refrac = total plasma protein measured by refractometry fTPP Col = total plasma protein measured by a colorimetric method

3 AST = aspartate transaminase h CK = creatine kinase iNa/K = Sodium/potassium ratio jTT4 radioimmunoassay = thyroxine measured by a radioimmonoassay k TT4 elisa = thyroxine enzyme linked immunoabsorbent assay lTT3 = triiodothyronine m = determined from whole blood

84 Table 4.1. Body weight, packed cell volume and plasma chemistry values from nestling marabou storks (n =20).

Parameter Mean SD Median Ql‘I Q3” Min Max Weight Kg 4.1 1.6 3.8 3.1 4.9 1.6 7.6 " PCV UL 0.34 0.04 0.33 0.31 0.37 0.28 040

° TPP Refrac g/L 46 4 47 43 49 40 52

'TPP Col g/L 41 5 42 39 45 30 50

”Glucose (mMol/L) 11.7 2.1 12.2 10.4 13.2 7.5 15.1

= AST U/L 135 28 143 115 155 84 178

Calcium mMol/L (N=18)° 2.70 0.17 2.74 2.68 2.82 2.35 2.95

Phosphorous mMollL 2.84 0.87 2.53 2.15 3.53 1.74 4.49

' CK U/L (N=16) 1422 575 1320 988 1854 584 2705

Uric Acid mMol/L 0.873 0.355 0.806 0.643 1.100 0.357 1.422

Albumin g/L 17.1 2.3 17.0 16.0 19.0 12.0 20.0

Globulin g/L 24.2 3.5 25.0 22.0 26.0 18.0 30.0

Sodium mMol/L (N=16)c 149 2 149 148 151 143 153

Potassium mMol/L 4.2 1.3 3.9 2.9 5.3 2.7 6.5

Chloride mMol/L (N=18)° 107 5 107 105 111 96 115 iNu/K ratio (N=16)c 37 12 35 27 48 23 54

Cholesterol mMol/L 4.97 1.27 5.10 3.86 5.98 3.08 7.28 j'1'r4 RIA nMol/l 8.9 9.0 6.7 2.0 14.2 0.0 20.0

" TT4 elisa nMol/l 15.98 18.05 8.27 7.18 22.65 2.00 28.60

“113 nMol/l (N=18)“ 1.61 1.55 0.41 1.33 1.88 0.90 2.30

85 Table 4.2: Results of Analysis of variance of plasma chemistry values in nestling

marabou storks (Leptoptilos crumeniferus) (n = 19) from Kampala, Uganda. a AST = aspartate transaminase b Chol. = cholesterol c CK = creatine kinase " TT4 ELISA = thyroxine enzyme linked immunoabsorbent assay

" TT4 RIA = thyroxine radioimmunoassay fTT3 = triiodothyronine

3 TPP (ref) = total plasma protein determined by reﬁactometry h TPP (col) = total plasma protein determined by the buiret method

’ BW = body weight

86 Table 4.2. Analysis of variance for plasma chemistry values in nestling marabou storks (n = 19)

o»- -—.~_.—-— -.—-- - .. .n.-.

Plasma Variables F P Plasma Variables F P

Chemistry Chemistry

aAST BW .12 .7284 Albumin BW 1.2 .2892

Gender 1.42 .2504 Gender 2.69 .1205

Overall .4779 Overall 1.95 .1752

cCK BW 10.45 .0052 Calcium BW .25 .6250

Gender .08 .7870 Gender .75 .3982

Overall 5.6 .0175 Overall .50 .6152

I'Chol. BW .22 .6427 Chloride BW .13 .7251

Gender 3.42 .0831 Gender 1.95 .1820

Overall 1.82 .1941 Overall 1.04 .3769

Globulin BW 6.47 .0217 Glucose BW 10.4 .0050

Gender 5.98 .0264 Gender .93 .3496

Overall 6.22 .0100 Overall 5.67 .0130

87 Table 4.2. (continued)

-.....— ... -—--e--—- m ml-w— - Plasma Variables 17 Variables F I

Chemistry Chemistry

BW .06 .8055 BW l .34 .2645

Potassium Gender . 12 .7358 Sodium Gender 2.23 .1551

Overall .09 .9141 Overall 1.78 .2002

BW .01 .9267 BW 2. 16 .1601

Phosphorous Gender . 16 .6914 PCV Gender 1 .39 .2539

Overall .09 .9180 Overall 1.7 8 .1993

BW 1 .25 .2806 BW .28 .6029

Gender 1 .08 .3133 °TT4 Gender .01 .9307

Overall 1.17 .3369 Overall .14 .8664

BW .09 .7655 BW .08 .7789

‘1'r4 ELISA Gender .02 .8869 Uric Acid Gender .01 .9278

Overall .06 .9453 Overall .04 .9561

BW 2.74 .1161 BW ' 3 .99 .0629

'TPP (Ref) Gender 9.53 .0067 “TPP Gender 4.82 .0432 (Col) Overall 6.14 Overall 4.41 .0298

88 Chapter 5

Persistent Organic Pollutant

and Mercury Concentrations in African Fish Eagles (Haliaeetus vocifer), Marabou Sto rks (Leptoptilos crumeniferus) and Tila pia

(Oreochromis niloticus) in Uganda

89 Abstract

The purpose of this research was to determine whether there are Signiﬁcant

differences in persistent organic pollutant (POP) and mercury concentrations in tissues

of African ﬁsh eagles (Haliaeetus vocifer) from Lake Victoria near Entebbe and Lake

Mburo. Secondly, we sought to determine POP and mercury concentrations in ﬁsh

from these two lakes in addition to the Nile river delta near Wanseko (Murchison Falls).

Thirdly, we sought to determine POP and mercury concentrations in marabou stork

(Leptoptilos crumeniferus) nestlings from urban Kampala. Total mercury was measured

in the breast feathers of 33 nestling and adult African ﬁsh eagles from Lake Mburo, and

Lake Victoria near Entebbe and 20 nestling marabou storks ﬁom Kampala, Uganda

from June 2002 through January 2003. Plasma concentrations of DDT, aldrin,

hexachlorocyclohexane (or-HCH), dieldrin, endrin, heptachlor and their metabolites, B-

HCH, 2,4’-DDD, 4,4’-DDD, 4,4’-DDE, 2,4’-DDT, 4,4’-DDT, heptachlor epoxide,

lindane, nonachlor and total polychlorinated biphenyl concentrations (PCBS) were also

measured. Total body burdens of these chemicals and mercury were measured in 18

Oreochromis niloticus, a ﬁsh eagle prey item at Lake Mburo, Murchison Falls and Lake

Victoria. Mercury concentrations were signiﬁcamly higher in eagle adults and nestlings

from Entebbe than adults and nestlings from Lake Mburo (p s 0.05). There were no

signiﬁcant differences (p s 0.05) in mercury concentrations between genders. Mercury

concentrations for marabou Stork nestlings in Kampala were slightly higher than

concentrations in ﬁsh eagle nestlings from Entebbe. However, there was no signiﬁcant

difference (p _<_ 0.05) in mercury concentrations between the entire ﬁsh eagle population

sampled at Entebbe and marabou stork nestlings sampled at Kampala. Five adult eagles

90 and five Oreochromis niloticus fish samples fi'om Entebbe had 4,4'-DDE in plasma in

the range 0.001 — 0.005 ppm. Marabou Stork samples did not contain 4,4'-DDE at 0.001

ppm. No samples contained PCBS at the 0.003 ppm detection level or other POPS

(besides 4,4'-DDE) at the 0.001 ppm limit of detection.

Introduction

The African ﬁsh eagle (Haliaeetus vocifer) is a widespread, often locally abundant tertiary avian predator in lake-based food chains throughout sub-Saharan

Africa. The Aﬁican Fish Eagle is very sensitive to the effects of persistent organic pollutants (POPS) (Douthwaite 1992). In this regard it is similar to other piscivorous tertiary avian predators such as the bald eagle (Haliaeetus leucocephalus) in North

America (Bowerman 1993) and the white tailed sea eagle (Haliaeetius albicella) in

Sweden (Helander et a1 1982).

The marabou stork (Leptoptilos crumeniferus) is resident in tropical Aﬁica where it is common to abundant in most parts of its range (Hancock et a1 1992). The marabou stork has responded to increasing urbanization and centralization of human populations by adapting a scavenging lifestyle in urban areas (Pomeroy 1977, 1978). Populations have increased in Kampala and breeding colonies may be found in the city center.

Although the marabou stork undertakes short seasonal migrations generally between the north and south of Uganda, 8 year round resident population in Kampala may be increasing (Hancock et al 1992). This and their geographical exposure in urban areas to greater industrial output may make these populations useful as indicators of exposure to a range of pollutants, including heavy metals such as mercury.

91 Studies on POP and metal contaminant concentrations in African ﬁsh eagles and marabou storks are limited and in the former Species, were largely done in southern

Africa (Davies and Randall 1989; Tannock et al 1983). South Africa and Zimbabwe had greater and more systemic use of POPS than Uganda during the latter decades of the twentieth century. However, the lack of accurate reporting procedures makes it difficult to determine the actual usage of many chemicals in Uganda. Ejobi et al (1996 a, b) quotes 80 tonnes of DDT per year being used mainly for cotton growing and mosquito control. Dieldrin was used at the rate of 392 tonnes per year for banana weevils and termites while 30 tonnes were also used for tsetse ﬂy control. Lindane, aldrin, hexachlorobenzene, campheclor, chlordane and heptachlor were also used. In comparison importations to South Africa between 1974 and 1976 averaged approximately 950 tonnes per annum and in 1982 Zimbabwe imported approximately 1000 tonnes (Davies &

Randall 1989). In 1985 DDT was banned for agricultural use in Zimbabwe. However usage was still reported to be up to 300 tonnes (Hartley and Douthwaite 1994) for

Zimbabwe and 121 tonnes for South Africa in the mid 19808 (Davies & Randall 1989).

No known studies on POPS in avian species have been conducted in Uganda.

Mercury pollution is a result of natural and anthropogenic activities. Natural degassing of the earth’s crust is the major source of environmental mercury worldwide

(Heinz 1996). Anthropogenic sources include industrial pollution and burning of fossil fuels which commonly occurs in the industrial area along the shore of Lake Victoria near

Entebbe and Kampala. Methylmercury, the most toxic form, can have halmﬁrl effects on

92 adult and ﬂedgling survival as well as reproduction, behavior and cellular development in avian species (Burger 1994). The neurotoxic effects of methylmercury can alter nesting behaviour and negatively impact the reproductive success of avian species.

Methylmercury can also speciﬁcally impair hatching. Methylmercury pollution is of most concern because it is biomagniﬁed along the food chain (Heinz 1996). A study has examined metal concentrations (zinc, cadmium, lead, copper, iron, manganese, chromium and cobalt) in the feathers of adult marabou storks from Kampala city and surrounding areas (Nyangababo 2003). This study did not include mercury. No studies have examined the potential impact of mercury pollution on wildlife in Uganda.

With increased political and economic stability, a rapidly expanding urbanized population, and industrial growth of 7 % (Central Intelligence Agency, 2002) baseline assessment and monitoring of pollution in Ugandan is needed. The bald eagle has been proposed as an ecosystem monitor species of North American Great Lakes water quality by the International Joint Commission (1985) and the 1998 State of the Lakes Ecosystem

Conference (SOLEC 1998). It was proposed as an ecosystem monitor particularly in regard to the toxic effects of organochlorine compounds on piscivorous wildlife

(Bowerman 2000). Given the African ﬁsh eagle’s widespread distribution and its relative abundance, it may also be a valuable indicator species of water quality in the sub-Saharan lake-based ecosystems of Africa, such as Lake Victoria.

93 The hypothesis to be tested by this research is no difference exists in persistent organic pollutant and mercury concentrations in tissues of African ﬁsh eagles from Lake

Victoria near Entebbe compared to those from Lake Mburo.

The objectives of this research were to establish baseline concentrations of POPS and total mercury in African ﬁsh eagle adults and nestlings at a site with signiﬁcant anthropogenic disturbance (Lake Victoria near Entebbe) and at a relatively undisturbed

Site (Lake Mburo). Whole body concentrations of POPS and total mercury were deterrrrined for Tilapia ﬁsh (Oreochromis niloticus), an eagle prey item, to demonstrate bioaccumulation of these pollutants within the aquatic food chain. As ﬁsh can virtually be the primary source of dietary protein for human populations in this region, sampling of

ﬁsh may also give some indication of human exposure to these pollutants. Marabou stork nestlings were sampled in Kampala as a contrast and comparison to the results from ﬁsh eagles, and to determine baseline concentrations of the pollutants in this common species.

A further objective of this study was to assess the feasibility of using ﬁsh eagles and marabou storks as biomonitors of environmental health. Multiple criteria including knowledge of species biology, breeding cycle, physiology, body size, diet and ease of sample collection determine species suitability for the role of biomonitor. To this end other baseline information on both avian species was collected including hematological, plasma chemistry values, body measurements and body weights. Adult ﬁsh eagles were banded and eagle nest site habitat preferences were also characterized.

94 The value of this study is in its establishment of baseline concentrations of contaminants in species that may ﬁrlﬁll the role of biomonitors of environmental health.

In addition, the study also developed logistical and technical methodologies for sampling these species. To maximize beneﬁts from this work, an assessment of the suitability of

African ﬁsh eagles and marabou storks as biomonitors of environmental health in African lake-based ecosystems was made. However, also required is further long - term monitoring of contaminant trends with research on population dynamics and water quality, together with the application of developed techniques to monitoring other pollutants and indices of environmental change. Other indices of environmental change include species distribution, abundance, community structure and breeding. Apart from changes associated with speciﬁc pollutants these indices may change with natural climactic variation, global warming, ozone depletion, habitat change and fragmentation and changes in species composition within an ecosystem (Jarvis 1993). Whatever the potential sources of change, this study establishes an important database of protocols and values to help measure both the change and its impacts on both human and wildlife populations.

Materials and Methods

African ﬁsh eagles were sampled at Lake Mburo, a 6 km long freshwater lake in

South Western Uganda (0°. 39' S, 30°. 57' E) situated in a 256 km2 national park, and on

Lake Victoria at Entebbe, from Nfo Island (0°. 00' N, 32°. 26' E) to Kisubi Bay (0°. 05' N,

32°. 35' E). Lake Mburo is 230 km and Entebbe 40 km from the capital Kampala. Fish eagles were sampled at Lake Mburo in July, August and December 2002 and at Lake

Victoria in August 2002 and January 2003. Thirty-three eagles were sampled: ten adults

95 and eight nestlings from Lake Mburo and ﬁve adults and ten nestlings from Lake

Victoria.

Twenty marabou stork nestlings were sampled from a colony located along Nile

Avenue, in central Kampala, Uganda (0°. 32' N, 32°. 58' E). Marabou Stork nestlings were sampled from twelve nests, in six trees (Tabebuia pentaphylla), in one colony, on

one day from 0700 till 1800, in January 2003.

Oreochromis niloticus were sampled at Lake Mburo and Lake Victoria at Entebbe

in the same locations and times that ﬁsh eagles were sampled. In addition ﬁsh were

sampled ﬁom the Nile river at Wanseko, close to Murchison Falls national park (2°. 15'

N, 31°. 38' E), in January 2002.

Adult ﬁsh eagles were captured on water using a ﬁsh “snare vest” technique.

Oreochromis niloticus were ﬁtted with eight to twelve 5 - 6 cm diameter ﬁshing line

snares (loops). The free ends of the snares penetrated the body of the ﬁsh and were then

tied on themselves and the excess line cut. The line was attached to a hand held wooden

reel and the snared ﬁsh was then placed in the water. Once captured, the ﬁeld crew

paddled to the eagle and retrieved it by securing both legs. Fish eagles swim well so there

was little risk of drowning. On shore, the eagles were placed in dorsal recumbency and

the eyes covered by a baseball cap. Ten ml of blood were collected from the brachialis

vein via a 21 or 23g x 1.90 cm (3/4 inch) butterﬂy catheter (Surﬂo Winged Infusion Set,

Terumo Medical Corporation, Elkton, MD) connected to a 10ml Sterile syringe and were

96 immediately transferred to a 10 ml lithium heparin vacutainer (Becton Dickinson,

Franklin Lakes, NJ). Four ml of additional blood were collected and transferred to a ﬁve m1 calcium EDTA vacutainer. Three blood smears were made from fresh blood using the slide on Slide technique (Campbell 1988). Fresh whole blood was used to determine blood glucose (Medisense 2Cm card glucometer, Medisense Inc. Bedford, MA). A drop of whole blood was placed on a commercially prepared paper sample card for molecular sex determination based on total erythrocyte DNA (Avian Biotech International, Tallahassee,

FL). Five whole breast feathers were hand plucked for determination of total mercury concentrations. A physical examination and visual description of any abnormalities was made. Hallux, culmen, footpad and eighth primary length, as well as bill depth were measured (Figure 3.1 and Figure 3.2). Body measurements methods used were those described for the bald eagle (Bortolotti 1984 a, b) (Figure 3.1 and 3.2). Birds were banded with 18-22mm internal diameter metal rivet bands inscribed with a three letter

sequential code and the word “MAKERERE” (Gey Band and Tag Company, Norristown,

Pennsylvania, USA). The bands were colored either red or gold for Lake Mburo and black for Entebbe. Suspected female birds were handed on the left leg and suspected

males on the right leg. Lastly, birds were placed in a cotton sack and weighed. (Horns

model 20 spring balance, Douglas Horns Corp. Belmont, CA). Eagles were then released

from land at the closest point possible to the capture location. Average time from capture

to release was 34 minutes (range 20-45 minutes).

African ﬁsh eagle and marabou stork nestlings were retrieved from the nest using

professional tree climbing techniques (USDA Forest Service 1996), with modiﬁcations

97 for tropical tree species and environmental conditions. The main method of tree ascent was using tree climbers (Klein Tools, Chicago ILL). Eagle and stork nestlings were

placed singly into a ventilated bag and lowered to the ground for sampling. Sampling of

nestlings was as described for adults except the volume of blood collected was between

four and 14 ml. No more than one percent of body weight was collected and oﬁen

considerably less. Venipuncture site on stork nestlings was the medial metatarsal vein.

Eagle’s ages were approximated (+/- 3 days) based on body weight by the calculations

presented in Sumba (1988). Average time from removal of the nestling until return was

14.5 minutes (range 7-22) for marabou storks and 32 minutes (range 10-60) for ﬁsh eagle

nestlings.

Oreochromis niloticus were sampled as they appeared to be the ﬁsh eagle’s main

prey item at Lake Mburo and Entebbe. Fish eagles feed on a variety of ﬁsh with cichlid

species, (including Oreochromis niloticus, that are herbivorous) occupying the littoral

zone comprising 80% of the ﬁsh taken (Stewart et al 1997). Fish samples were '

purchased from local ﬁsherman as close as possible to where eagles were sampled. Fish

less than 300-400 g and 300-450 mm in length were sampled as this closely corresponded

to the limited information available on ﬁsh eagle prey selectivity (Stewart et al 1997).

Fish were weighed and their body length recorded. Mean length and weight for ﬁsh from

Lake Mburo were 225 mm and 343 g, Entebbe 267 mm and 569g and Murchison Falls

253 mm and 411g. A 100g cross trunk sample of ﬁsh including skin, muscle, bone and

viscera was dissected just cranial to the dorsal ﬁn and weighed. Samples were placed in

freezer bags and transferred to a MVE Doble-20 dual-purpose vapor shipper/liquid

nitrogen tank (MVE Bio-Medical Systems, Bumsville, MN).

98 Samples were placed in a chilled cooler after collection. Time of sampling to storage of plasma in liquid nitrogen was 3.5 hours (range 2-9 hours) for eagles and 8 hours (range 4-16 hours) for marabou storks. Samples were centrifuged for ten minutes at

3000 rpm (vulcon mobilespin P8126-6, vulcon technologies, Grandview, MO). Plasma was removed and divided into ﬁve 2 ml cryovials per bird (Cryogenic Vial, Corning Inc.

Corning, NY) that were deposited into the vapor shipper/liquid nitrogen tank. Plasma samples were transported to the Diagnostic Center for Population and Animal Health

(DCPAH) at Michigan State University’s Veterinary Medical Center (MSU) then transferred to a - 80 °C ultra-low freezer until analyzed. Analysis occurred within two months of sampling in all cases.

All analysis was performed at the DCPAH at the MSU toxicology laboratory.

Feathers for total mercury analysis were washed in reagent grade acetone, deionized water, then again in chromatography grade acetone (Burdick & Jackson, Muskegon, MI) to remove debris. Feathers were allowed to dry overnight in a fume hood and then weighed. Feathers were then placed in sealed 30 ml teflon vessels in a concentrated 2 ml nitric acid digest (Instra-analyzed grade, J .T. Baker Inc, Phillipsburg, NJ) at 95°C overnight. Ten ml of the digest was quantitatively transferred to a volumetric flask and mixed. A final acid concentration of 7 % HCl acid and 2 % HNO3 to match the standard curve solutions 0, 25, 100, 500 ppt were made and diluted to volume. Aliquots were taken from the ten ml flask and diluted based on feather weight to achieve total feather dissolution and analyzed by cold vapor atomic absorption spectrophotometry at 253.7 nm. (LCD mercury monitor 3200, Thermo Separation Products, Riviera Beach, FL).

99 Commercial procedural blanks (2976 Mussell tissue with concentration 53.8 ppb to 68.2 ppb. National Institute of Standards and Technology Gaithersburg, MD) were used to monitor the accuracy of the analysis. Fish tissues were snap frozen for 60 seconds in liquid nitrogen then milled in an AIO analytical mill (Janke & Kunkel, IKA Labortechnik

Germany). One gram of powdered ﬁsh was added to 2 ml of concentrated nitric acid.

Further analysis followed the procedures outlined for feather sampling.

Tissues (serum, ﬁsh) for analysis of POPS were extracted and puriﬁed by the procedure described by Price et al (1986). Concentrations were determined by gas chromatography. Equipment used was a Varian 3400 gas chromatograph (capillary column method) with electron capture detector.

Analysis of variance (ANOVA) was conducted to assess the association between site, age (nestling or adult), gender and breast feather mercury concentrations in ﬁsh eagles and between body weight, gender and mercury concentrations in marabou storks

(SAS PROC ANOVA for categorical risk factors, and SAS PROC GLM for continuous risk factors. SAS 8.2, 2001.SAS Inc., Cary, NC). Analysis of variance was also used to assess the associations between site, body weight and total body mercury burdens in

Oreochromis niloticus. These analyses were conducted both at the univariable (only one risk factor at a time) and multivariable level. Multivariable analyses were conducted to adjust the effect of selected risk factors simultaneously. The level of signiﬁcance for type

1 (a) error was set at a probability of 0.05. Descriptive statistics were done using Excel

(Microsoft Excel, Microsoft Ofﬁce 2000 Professional. Microsoft Corporation, Redmond,

WA). Descriptive statistics are emphasized due to the small sample size. This emulates

100 the methodology of other studies examining wild avian species where only small sample

Sizes could be obtained (Garcia- Montijano 2002; Lumsden 1998).

This study was conducted under approval of the Michigan State University All

University Committee on Animal Use and Care. The Uganda National Council for

Science and Technology and the Uganda Wildlife Authority granted research permits for

this project.

Results

Total mercury concentrations in breast feathers from adult and nestling Aﬁican

ﬁsh eagles, nestling marabou storks and in whole body section samples ﬁ'om

Oreochromis niloticus are presented in Table 5.1. Feather mercury concentrations are

presented on a dry weight basis and ﬁsh sample concentrations are presented on a wet

weight basis. Analysis of variance for total mercury concentrations in Aﬁican ﬁsh eagle

and marabou stork breast feathers are presented in tables 5.2 and 5.4 respectively.

Analysis of variance comparing mercury concentrations between ﬁsh eagles from Lake

Victoria near Entebbe and marabou storks ﬁom Kampala are presented in table 5.3.

Analysis of variance assessing the association between site, body weight and mercury

concentrations in Oreochromis niloticus is presented in table 5.5.

Mercury concentrations in adult ﬁsh eagles from Entebbe were Signiﬁcantly

higher than concentrations in adults from Lake Mburo (p s 0.05). Similarly, mercury

concentrations in nestling ﬁsh eagles from Entebbe were signiﬁcantly higher than

101 concentrations in nestlings fi’om Lake Mburo (p s 0.05) (Figure 5.1). There were no significant differences (p 2 0.05) in mercury concentrations in breast feathers of fish eagles based on gender or whether the bird was an adult or fledgling. Marabou stork nestlings from Kampala had slightly higher concentrations of mercury than fish eagle nestlings from Entebbe (Entebbe is 40 km from Kampala). However, there was no significant difference (p s 0.05) in mercury concentrations between the fish eagle populations from Entebbe and the nestling marabou stork population from Kampala.

Two adult male ﬁsh eagles from Entebbe had much higher mercury concentrations (2.3 ppm and 1.4 ppm wet weight) than other ﬁsh eagles sampled.

There was a significant difference in mercury concentrations in whole body cross section samples of Oreochromis niloticus fi'om the three study sites (p _<. 0.05). The highest concentration of mercury in fish came from the Murchison Falls site.

African fish eagle plasma, marabou stork plasma and whole body section samples from Oreochromis niloticus did not contain concentrations of the following chemicals at the 0.001 ppm limit of dedetection: aldrin, DDT, a-BHC, dieldrin, endrin, heptachlor and their metabolites, B-BHC, 2,4’-DDD, 4,4’-DDD, 2,4’-DDT, 4,4’-DDT, heptachlor epoxide and lindane and nonachlor. Total PCBS were not detected in Afiican fish eagle plasma, marabou stork plasma or whole body section samples from Oreochromis niloiicus at the detection limit of 0.003 ppm. Five adult eagles from Entebbe had 4,4'-

DDE detectable in plasma, one at 0.005ppm (male), one at 0.003ppm (female), two at

0.002ppm (males) and one at 0.001ppm (female) wet weight. Samples were not

102 corrected for lipid content of plasma. Five Oreochromis niloticus samples from Entebbe contained 4,4'-DDE concentrations of 0.001, 0.001, 0.002, 0.003 and 0.003 ppm wet weight. Six eagle nestlings from Entebbe and one adult from Lake Mburo tested positive for 4,4'-DDE but at concentrations lower than 0.001ppm. At 0.001 ppm 4,4'-DDE was not detected in marabou stork plasma and whole body section samples of Oreochromis niloticus.

Discussion

Mercury

Scheuharnmer and Bond (1991) suggest that mercury feather concentrations greater than 20 ppm may be associated with toxic effects in birds. Eisler (1987) states that concentrations above 5ppm fresh weight are thought to be associated with adverse affects in sensitive avian species. However, this study was not based on piscivorous birds but grain eating pheasants and mallards. Risk categories for mercury accumulation in common loons (Gavia immer) use 0-9 ppm wet weight as a low risk category that indicates background mercury concentrations in loans that have been minimally impacted by human activity (Schoch & Evers 2002). Jagoe et al (2002) found feather mercury concentrations in South Carolina bald eagle nestlings to be 3.06 ppm dry weight. They did not make any conclusions as to the affects of this level of exposure. Bowerman

(1993) found concentrations in adult body feathers and nestling bald eagle feathers from the Great Lakes basin of North America of 21.4 ppm and 9.0 ppm respectively. He concluded that neither productivity (young per occupied nest) nor success was

103 Signiﬁcantly (p s 0.05) correlated with logarithmic concentrations of adult or nestling feather mercury. It may also be hard to determine the affects of mercury on reproduction when there are also tissue concentrations of chemicals such as DDE (Bowerman 1993).

Mercury concentrations found in all age groups of African ﬁsh eagles and nestling marabou storks were well below all of the concentrations mentioned above.

Mercury accumulates over time with the two main routes of excretion being growing feathers and eggs (Heinz 1996). Therefore it would be expected that mercury concentrations would be greater in adult birds than nestlings and greater in male adults than female adults. Females would be expected to excrete mercury into the egg and also have a larger body weight than males and thus would be expected to have lower concentrations than males. The results of this study do not support these expectations, as there was no significant (p _<_ 0.05) difference in mercury levels based on age (nestling or adult) or gender. When Burger (1994) summarized studies examining gender differences in mercury concentrations in feathers, she found only two out of eight studies that found significant differences. She also reported that most studies have failed to find any relationship between mercury levels and age. Although these studies may support our

ﬁndings, the results may also be due to a small sample size. This reinforces the fact that due to the small sample size the results Should be interpreted with caution. Sampling occurred at times of egg laying and at times of non breeding and it was impossible to determine the reproductive status of most adult female birds sampled. It was therefore impossible to determine whether egg laying, at some point in time before new feather growth, may have lowered the total body mercury concentrations through excretion of mercury into the egg. Another factor in this study was that a small percentage of the

104 samples ﬁ'om very small nestlings were breast down rather than breast feathers. Gariboldi et a1 (2001) found a signiﬁcant (p S 0.05) correlation between mercury concentrations in blood, down and feathers of wood storks (Mycteria Americana) (n = 300) at four out of

ﬁve sites in the southeast USA. There is some debate over whether the mercury in down originates mainly from the egg (Becker et a1 1994) or dietary accumulation (Gariboldi

2001). This could be variable between species. The higher mercury concentrations in fish eagles from Entebbe may reflect greater bioaccumulation through dietary exposure than occurred with the fish eagles at Lake Mburo. Oreochromis niloticus, one of the main fish eagle prey items at both sites, had higher concentrations in fish from Entebbe than those from Lake Mburo. We do not believe the results for the two adult fish eagles with relatively high mercury concentrations compared to all other fish eagles sampled are spurious as they were captured in close proximity to each other, were nesting on the property of a large flower factory and both had 4,4'-DDE detectable at 0.002 ppm wet weight in plasma. Again, the results should be interpreted cautiously as the findings require more investigation given the small sample size.

Anthropogenic mercury emissions associated with fossil ﬁrel burning are a possible explanation for the higher mercury concentrations in feathers returned from

Entebbe and Kampala samples compared with samples from Lake Mburo. The fact the human population of Kampala increased from 774,241 in 1991 to a preliminary ﬁgure of

1,208,544 in 2001 and use of diesel fuel (inferior grades and quality) in Uganda increased from 125,621 in 1997 to 207,183 cubic metres in 2001 (Uganda Bureau of Statistics,

2002) lend support to this explanation. Another source of mercury may be point source emissions through the burning of garbage, which is a widespread practice in the

105 communities around the Lake Victoria shoreline. A factor that may have contributed to natural background mercury concentrations at all sites was the eruption of two volcanoes,

Nyamuragira and Nyaragongo in the Democratic Republic of the Congo, near the border with Uganda, in July 2002. Volcanic eruptions are a natural source of atmospheric mercury, as is degassing of the earths crust (Heinz 1996).

Adult African fish eagles are extremely territorial (Brown 1980) thus concentrations of mercury in nestlings would be expected to reflect environmental concentrations in and around the nest site and confluent fishing territory. Therefore contaminant concentrations in nestlings and adults would be expected to reflect local concentrations of contamination. The significantly (p s 0.05) higher concentration of mercury in fish from Murchison Falls compared with Lake Mburo or Entebbe (p = 0.02) may reflect sources of contamination upstream from this sampling site or local conditions that favor the mobilization of mercury. Further investigation of the complex ecology of the area, the performance of water quality analysis and samples from fish eagles would be required to verify these assertions.

Feathers are an excellent, non-invasive tissue to sample for mercury analysis.

Feathers contain higher concentrations of mercury than in other body organs

(Westermark et al 1975) with approximately 70% of total body mercury found in feathers. (Honda et al 1986). Feather mercury concentrations can easily be measured. A positive correlation has also been shown between mercury content in feathers and internal tissues (Thompson et a1 1991). Mercury concentrations are spread evenly throughout the feather, indicating endogenous incorporation methods, rather than atmospheric deposition

106 (Hahn et a1 1993). Mercury concentrations of ﬁsh eagles and marabou storks can therefore be directly associated with dietary exposure. This dietary exposure may occur through bioaccumulation of mercury within the food chain. Feathers are a major excretory pathway for body mercury during molt as concentrations accumulate during periods of feather growth. Differing mercury concentrations are found depending on the order the sampled feather has in the sequence of the molt cycle. New feathers grown early in the cycle would be expected to have higher mercury levels than those grown later in the cycle. As the molt cycle continues and more mercury is excreted into growing feathers the total body pool of mercury is diluted. Thus the later in the molt cycle a new feather is grown the less mercury will be found in the feather. This has not been consistently shown to occur in all individuals within a population. Body contour feathers are the feather type that shows the least variation in mercury content (Fumess et al 1986).

This supports the methodology in this project of taking a pooled sample of breast feathers. Interpretation of total feather mercury concentrations should be made with caution. For example trace metals like selenium and zinc can lower the affects of high tissue residue concentrations of mercury. In addition, certain marine mammals and birds can demethylate methyl-mercury into inorganic mercury (Heinz 1996). It is suggested that similar mechanisms to tolerate high mercury concentrations may exist inpiscivorous raptor species (Norheim & F roslic 1978). However, 95% of total mercury found in blood, feathers and eggs is in the methyl-mercury form (Thompson, 1996) and therefore total mercury concentrations may be a reasonable indicator of the potential for adverse affects in avian species.

107 Mercury concentrations in marabou stork nestling feathers were well below the concentrations reported to cause adverse reproductive and physiological affects in other avian species (Scheuhammer and Bond 1991; Eisler 1987; Scoch and Evers 2000).

However, Gariboldi et al (2001) point out the need to relate tissue concentrations in piscivorous birds to sublethal effects at both the individual and population level. This would also apply to marabou storks and African fish eagles. The trend towards higher feather mercury concentrations in marabou stork nestlings compared to fish eagle nestlings from the same region (Entebbe and Kampala are 40 km from each other) may indicate species differences in accumulation or greater dietary exposure in marabou storks. The lack of any significant difference (p 2 0.05) in mercury concentrations between the fish eagle population from Entebbe (including adults), and the nestling marabou stork population from Kampala may be caused by greater dietary exposure in marabou stork nestlings. This may be due to point source emissions such as the burning of garbage. Marabou storks also have a longer fledging period than fish eagles (135 days for marabou storks and 75 days for fish eagles) and therefore a longer period of feather growth and potential accumulation of mercury in the growing feather. The mercury concentrations found in marabou storks nestlings from Kampala may reinforce the conclusions in relation to concentrations of mercury found in fish eagles from the

Entebbe region. That is, the concentrations of mercury found in biota at

Entebbe/Kampala are higher than the concentrations found at Lake Mburo. In addition, this may reﬂect bioaccumulation from dietary sources. Marabou storks in Kampala have largely adapted to a cosmopolitan, broad ranging diet based on scavenging and it is stated that they will eat virtually anything organic (Hancock et al 1992). During nestling

108 growth, the parent marabou storks seek out dietary protein sources to cater for the extra grth requirements of nestlings (Hancock et al 1992). It is postulated that this may lead to seeking of food items that originate at higher trophic levels in the food chain, thus increasing susceptibility to persistent bioaccumulative endogenously derived toxicants.

Concentrations of mercury in marabou stork nestlings in Kampala may be seen as a non- speciﬁc indication of mercury concentrations contained in human organic reﬁise in

Kampala. However, caution should be exercised in attempting to ascribe exposure routes for mercury concentrations found in the marabou stork nestlings. Other exposure routes, in addition to diet may contribute to mercury concentrations in these birds. Atmospheric concentrations of mercury may also be higher in the center of a busy city and this may make a small contribution, through atmospheric deposition, and inhalation to the overall mercury concentrations found in the feathers of these marabou stork nestlings.

Persistent organic pollutants

Concentrations of 4,4'-DDE found in some adult ﬁsh eagle samples were below wet weight plasma concentrations linked to low productivity levels in bald eagle nestlings. Elliot & Norstrom (1998) calculated that 0.04 ppm wet weight DDE in plasma equals about 6 ppm DDE in eggs, and this level equates to a productivity of 0.52 young per occupied nest using the unweighted linear model of Wiemeyer et a1 (1993).

Productivity of 0.7 young per occupied territory is necessary for population maintenance in the bald eagle and reproduction is considered impaired when productivity

(young/occupied nest) is less than 1.0 (Bowerman 1993). The relationship between mean annual productivity and mean plasma 4,4'-DDE concentrations established for ten

109 subpopulations of bald eagles in the upper Midwest of the USA had productivity falling below 1.0 at approximately 0.011 ppm mean wet weight plasma 4,4'-DDE in nestlings

(Bowerman 1993). Depending on the species, DDE concentrations in eggs of 2-20 ppm wet weight are associated with eggshell thinning, breakage and breeding failure sufﬁcient to cause population declines (Hartley & Douthwaite 1994; Blus 1982; Henny et al 1984;

Weimeyer et al 1984). Bald eagle productivity has been found to be near normal at breeding areas where DDE was 3 ppm or less wet weight in eggs (Weimeyer et al 1984).

It must be remembered that all samples above the limits of detection in this study were adult birds and only trace levels below the 0.001ppm limit of detection were found in plasma from ﬁsh eagle nestlings. Unlike nestlings in the bald eagle studies, the adult birds in this study could not be accurately aged. Smith and Bouwman (2000) recorded a range of chlorinated hydrocarbon residues measured in blood plasma in four species of raptor ﬁ'om the North-West province of South Africa. Concentrations of DDE ranged between 0.0015 ppm to 0.0066 ppm (uncorrected wet weight) except for Lanner falcons

(Falco biannicus), an omiphagous species, had concentrations of 0.014 ppm. These levels were in adult birds and are similar to those recorded for the birds in this study. The authors of the South African study, which extrapolated egg DDE plasma concentrations from blood plasma, concluded that raptors are in no immediate danger of reproductive impairment on account of egg DDE residues. Evans and Bouwman (2000) also recorded blood plasma concentrations of DDE in pied kingfishers Ceiyle rudis from various areas of northern KwaZulu-Natal, South Afiica. Concentrations varied by location with the highest wet weight uncorrected concentration being 0.18 ppm. Pied kingfishers are common in Uganda and occur at a relatively high trophic level. It may be of interest to

110 sample these birds in the future. All the quantiﬁable concentrations of POPS in this study were in samples from the Entebbe area of Lake Victoria that also had higher mercury concentrations suggesting more effects of anthropogenic disturbance than at Lake Mburo.

Concentrations of POPS were quantiﬁed in human and cows milk in the Kampala region in 1992/1993 (Ejobi et al 1996 a,b). Pesticide residues detected were dieldrin, 4,4’-DDD,

4,4’-DDE, 2,4’-DDT, 4,4’-DDT, B-HCH, a-HCH and lindane. Mean 4,4’-DDE residues reported in human milk were 2.84 +/- 0.255 mg/kg milk fat and in cows milk 0.034 +/-

0.0004 mg/kg milk fat. These ﬁgures are not corrected for, percent recoveries. The mean percent extractable fat from cows milk was 2.9% and ﬁ'om human milk 4.3%. NO PCBS were detected in these studies. Cows in this study were usually not grazed, but fed on crop residues that could have been contaminated with pesticides. Concentrations of pesticides were generally lower than concentrations reported in other developing countries. The widespread, systematic usage of persistent organochlorinated pesticides that occurred in African countries, such as Kenya, South Africa and Zimbabwe in the

19708 and 1980s surpassed usage in Uganda. This was probably due to the lesser development of plantation agriculture in Uganda during its time as a British protectorate in addition to the prolonged period of civil unrest that continued from independence in the early 1960s till the mid 1980s

Raptor reproductive measures have been used to determine the effects of

environmental contamination on the health of raptor populations, such as the bald eagle

in North America. Bowerman (1993) in relation to bald eagles deﬁned productivity as

the number of young per occupied nest for each breeding area and success as the percent

of occupied breeding areas ﬂedging at least one young. Using these deﬁnitions and our

111 general observations, the results of the mercury/POP analyses would suggest healthy, self

sustaining populations of ﬁsh eagles at Lake Mburo and Entebbe. Out of ten nests

climbed that contained chicks only one had a single chick, seven had two chicks and two

had three chicks. Five other nests contained eggs, three with two eggs and two with three

eggs. Our general observations suggest that most of these nests were ﬂedging two chicks.

At Lake Mburo the average of three days of counting eagles from a boat led to a result of

74 adults and 15 immature ﬁsh eagles. The percentage of immature eagles was 20%. It is

recognized that no scientiﬁc methodology was employed in this count and it is presented

as observation only. Some authors suggest that a population of ﬁsh eagles with 20%

immature birds can be considered a reproductively healthy population (Brown and Cade

1972). To determine the effects of environmental contamination on a population requires

multiple sampling over many years as well as a thorough knowledge of the Species

biology and reproductive cycle at the study area. Again this was beyond the resources

available for this study but given the residue concentrations reported here, and our

general observations on productivity it could be inferred that environmental

contamination by these chemicals does not play a signiﬁcant role in determining the

population dynamics of African ﬁsh eagles at either of the sites studied. However, this

was not proven. We conclude that concentrations of mercury in African ﬁsh eagles is

unlikely to be a major factor in determining population dynamics of ﬁsh eagles at either

Lake Mburo or Lake Victoria at Entebbe. That there were no DDT residues found at the

limits of detection for ﬁsh eagles, marabou storks or Oreochromis niloticus sampled in

this study, may indicate that there is no ongoing exposure to this pesticide. It is worth

noting, however the similar concentrations of plasma DDE in the Entebbe ﬁsh eagle

112 samples to those from South Aﬁican raptors reported by Smith & Bouwman (2000).

South Africa is generally thought to have had greater systematic usage of DDT than occurred may have occurred in Uganda. The concentrations of DDE in African fish eagles at Lake Victoria at Entebbe may warrant further investigation. This should include a larger sample size, analysis of water quality and multi year examination of fish eagle productivity. We also conclude that mercury concentrations in fish eagles are significantly higher at the more urbanized Entebbe site than at Lake Mburo. This was supported by similar mercury concentrations being found in nestling marabou storks from

Kampala, which is almost contiguous with the Entebbe site. An expanding economy, population centralization, increasing use of fossil fuels and burning of garbage indicates that it would be reasonable to expect mercury concentrations to increase with time in wildlife living in the Entebbe/Kampala region.

113 References

Becker PH, Henning D & Furness RW. 1994. Differences in mercury contamination and elimination during feather development in gull and tern broods. Arch. Environ. Contam. Toxicol. 27: 162-167.

Blus. LJ. 1982. Further interpretation of the relation of organochlorine residues in brown pelican eggs to reproductive success. Environ. Pollut. (A) 28: 15-33.

Bortolotti GR. 1984 a. Criteria for determining age and sex of nestling bald eagles. J. Field Ornith. 55: 467-481.

Bortolotti GR. 1984 b. Sexual dimorphism and age related size variation in bald eagles. J. Wildl. Manag. 48: 72-81.

Bowerman WW. Regulation of bald eagle (Haliaeetus leucocephalus) productivity in the Great Lakes basin: an ecological and toxicological approach. PhD theses. Michigan State University, 1993. pp 159-185.

Bowerman WW, Best DA, Grubb TG, Sikarskie JG & Giesy JP. 2000. Assessment of environmental endocrine disruptors in bald eagles of the great lakes. Chemosphere. 41: 1569-1574.

Brown L. 1980 The African Fish Eagle. Bailey Bros & Swinfen Ltd. Folkestone, England.

Brown L H & Cade TJ. 1972. Age classes and population dynamics of the bateleur and African ﬁsh eagle. Ostrich. 43: 1-16

Burger J. 1994. Metals in avian feathers: bioindicators of environmental pollution. Rev. Environ. Toxicol. 5: 203-311.

114 Campbell TW. Avian Hematology and Cytology. 1988. Iowa State University Press. Ames. Iowa. Pp. 3-17.

Central Intelligence Agency. 2002. The World Factbook. Government Printing Ofﬁce, Washington DC.

Davies RAG & Randall RM. 1989. Historical and geographical patterns in eggshell thickness of African ﬁsh eagles Haliaeetus vocifer, in relation to pesticide use within southern Aﬁica. In: Raptors in the modern world. BU Meyburg & RD Chancellor (Eds) World Working Group on Birds of Prey and Owls. Berlin, London & Paris. pp. 501-513.

Douthwaite RJ. 1992. Effects of DDT on the ﬁsh eagle Haliaeetus vocifer population of lake Kariba in Zimbabwe. Ibis. 134: 250-258.

Eisler R. 1987. Mercury hazards to ﬁsh, wildlife and invertebrates: a synoptic review. US Fish and Wildlife Service Biological Report 85 (1.10), 90 pp

Ejobi F, Kanja LW, Kyule MN, Muller, P, Kruger, J. & Latigo, AAR. 1996 a. Organochlorine pesticide residues in mothers’ milk in Uganda. Bull. Environ. Contam. Toxicol. 56: 873-880.

Ejobi F, Kanja LW, Kyule MN, Muller P, Kruger J, Nyeko JPH & Latigo AAR. 1996 b. Organochlorine pesticide residues in cow’s milk in Uganda. Bull. Environ. Contam. Toxicol. 56: 551-557.

Elliott JE & Norstrom RJ. 1998. Chlorinated hydrocarbon contaminants and productivity of bald eagle populations on the paciﬁc coast of Canada. Environ. Toxicol. Chem. 17:1142—1153.

Evans SW & Bouwman H. 2000. The geographic variation and potential risk of DDT in the blood of pied kingﬁshers from northern KwaZulu-Natal, South Aﬁica. Ostrich. 71: 351-354.

Furness RW, Muirhead SJ & Woodbum DM. 1986. Using bird feathers to measure mercury in the environment: relationships between mercury content and moult. Mar. Pollut. Bull. 17: 27-30.

115 Gariboldi JC, Lawrence Bryan A Jr, & Jagoe CH. 2001. Annual and regional variation in mercury concentrations in wood stork nestlings. Environ. Toxicol. Chem 20: 1551- 1556.

Garcia-Montijano M A, Lemus JA, Montesinos A, Canales R, Luaces l & Pereira P. 2002. Blood chemistry, protein electrophoresis, and hematological values of captive Spanish imperial eagles (Aquila adalberti). J. Zoo & Wild. Med. 33: 112-117.

Hahn E, Hahn K & Stoeppler M. 1993. Bird feathers as bioindicators in areas of the German environmental specimen bank- bioaccumulation of mercury in food chains and exogenous deposition of atmospheric pollution with lead and cadmium. The Sc. of the Total Environ. 139/140: 259-270.

Hancock J A, Kushlan J A & Kahl MP. 1992. Storks, Ibises and Spoonbills of the World. Academic Press Lim. San Diego, California, USA. Pp. 133-138.

Hartley RR & Douthwaite RJ. 1994. Effects of DDT treatments applied for tsetse fly control on the Afiican goshawk in north-west Zimbabwe. Afi'. J. Ecol. 32: 256-272.

Heinz GH. 1996. Mercury poisoning in wildlife. In: Fairbrother A, Locke LN & GN Hoff. (Eds) Noninfectious Diseases of Wildlife. Iowa State University Press. Ames, Iowa pp. 118-127.

Henny CJ, Blus LJ, Krinitsky AJ & Bunck CM. 1984. Current impact of DDE on black crowned night-herons in the interrnountain west. J. Wildl. Manage. 48: 1-13.

Honda K, Nasu T & Tatsukawa R. 1986. Seasonal changes in mercury accumulation in the black eared kite, Milvus migrans lineatus. Environ. Pollut. (A) 42: 325-334.

International Joint Commission. 1985 Report of the Aquatic Ecosystem Objectives Committee to the Great Lakes Scientiﬁc Advisory Board. Windsor, Ontario.

116 Jagoe CH, Bryan AL Jr, Brant HA, Murphy TM & Brisbin IL Jr. 2002. Mercury in bald eagle nestlings from South Carolina, USA. J. of Wild. Dis. 38: 706-712.

Jarvis PJ. 1993. Environmental changes. In: Birds as monitors of environmental change. RW Fumess & JJD Greenwood (Eds) Chapman and Hall, London. Pp. 42-85.

Lumsden JH. 1998. “Normal” or reference values: questions and comments. Vet. Clin. Pathol. 27: 102-106.

Norheim G & Froslic A. 1978. The degree of methylation and organ distribution in some birds of prey in Norway. Acta Pharmacologia et Toxicologia. 43: 196-204.

Nyagababo JT . 2003. Trace metals in the contour feathers of marabou stork (Leptoptilos crumeniferus) from Kampala city and its surrounding areas. Bull. Environ. Contam. Toxicol. 70: 792-799.

Pomeroy DE. 1977. The biology of marabou storks in Uganda. 1. Some characteristics of the species, and the population structure. Ardea, 65: 1-24.

Pomeroy DE. 1978. The biology of marabou storks in Uganda. II. Breeding biology and general review. Ardea. 66: 1-23.

Price HA, Welch RL, Scheel RH, Warren LA. 1986. Modiﬁed multiresidue method for chlordane, toxaphene and polyclorinated biphenyls in ﬁsh. Environ. Contam. & Tox. 37: 1-9.

Scheuhammer AM & Bond D. 1991. Factors affecting the determination of total mercury in biological samples by continous-ﬂow cold vapor atomic absorption spectrophotometry. Biol. Trace Elem. Res. 31: 119-129.

Schoch N. & Evers DC. 2002. Monitoring mercury in common loons: New York ﬁeld report, 1998-2000. Report BRI 2000-01 submitted to US. Fish and Wildlife Service and New York State Dept. Environ. Conservation. Biodiversity Research Institute, Falmouth, Maine. pp. 1-18.

117 Smith I & Bouwman H. 2000. Levels of organochlorine pesticides in raptors from the north west province, South Aﬁica. Ostrich. 71: 36-39.

SOLEC. 1998. Selection of indicators. State of the Lakes Ecosystem Conference 1998. N. Stadler-Salt and P. Bertram (Eds) Hamilton, Ontario, Canada.

Stewart K M, Matthiesen DP, Leblanc L & West J. 1997. Prey diversity and selectivity of the African ﬁsh eagle: data from a roost in northern Kenya. Afr. J. of Ecol. 35:133-145.

Sumba SJA. 1988. Nestling growth in the African ﬁsh eagle in Uganda. Afr. J. of Ecol. 26: 315-321.

Tannock, J., Howells, W.W. and Phelps, RJ. 1983. Chlorinated hydrocarbon pesticide residues in eggs of some birds in Zimbabwe. Environmental Pollution (Series B) 5: 147- 155.

Thompson DR Hamer KC & Fumess RW. 1991. Mercury accumulation in great skuas Catharacta skua of known age and sex and its effects upon breeding and survival. Journal of Applied Ecology. 28: 672-684.

Thompson DR 1996. Mercury in birds and terrestrial mammals. In: WN Beyer, GH Heinz & AW Redmon-Norwood. Environmental Contaminants in Wildlife: Interpreting Tissue Concentration. Lewis Publishers. Boca Raton, Florida. pp. 341-356.

Uganda Bureau of Statistics. 2002. Key Economic Indicators. 47‘“ Issue. Ch 39. Uganda Bureau of Statistics, Entebbe, Uganda.

United States Department of Agriculture Forestry Service. 1996. National tree climbing ﬁeld guide, USDA Forest Service. Missoula Technology and Development Center, Missoula, Montana, 1996.

Wiemeyer SN, Bunck CM & Stafford CJ. 1993. Environmental contaminants in bald eagle eggs-1980-1984-and further interpretations of relationships to productivity and shell thickness. Arch. Environ. Contam. Toxicol. 24: 213-227.

118 Wiemeyer SN, Lamont TG, Bunck CM, Sindelar CR, Gramlich FJ, Fraser JD & Bird MA. 1984. Organochlorine pesticide, polychlorobiphenyl, and mercury residues in bald eagle eggs-1969-1979-and their relationship to eggshell thinning and reproduction. Bull. Environ. Contam. Toxicol. 13: 529-549.

Westermark T, Odsjo T & Johnels AG. 1975. Mercury content of bird feathers before and after Swedish ban on alkyl mercury in agriculture. Ambio. 4: 87-92.

119 Table 5.1 Total mercury concentrations (ppm dry weight) in breast feathers from adult and nestling African ﬁsh eagles (Haliaeetus vocifer) and nestling marabou storks

(Leptoptilos crumeniferus) and in whole body section samples of tilapia (Oreochromis niloiicus) (ppm wet weight) from Uganda.

" Q1 = 25th percentile of the sample size b Q3 = 75'“ percentile of the sample size

“ number indicates sample size

" LM= Lake Mburo

° Ebb= Lake Victoria near Entebbe

120 Table 5.1 Mercury concentrations in marabou storks, African ﬁsh eagles and Tilapia (Oreoclrromis niloticus).

Mean Median SD Q1' Q3“ Min Max

FISH EAGLES (feathers ppm dry weight) Adult LM" (10)° 0.357 0.314 0.124 0.249 0.441 0.230 0.554

Adult Ebbe (5) 1.059 0.588 0.787 0.548 1.390 0.471 2.300

Adult Male (9) 0.730 0.450 0.680 0.320 0.590 0.230 2.300

Adult Female (6) 0.384 0.378 0.150 0.249 0.519 0.23 0.548

Adult Total (15) 0.591 0.450 0.552 0.299 0.551 0.230 2.300

Nestling LM (8) 0.185 0.171 0.100 0.110 0.210 0.096 0.398

Nestling Ebb (10) 0.666 0.669 0.230 0.489 0.840 0.309 1.03

Nestling Total (18) 0.452 0.424 0.305 0.205 0.702 0.096 1.03

Eagle Total (33) 0.515 0.450 0.433 0.233 0.588 0.096 2.300

STORKS (feathers ppm dry weight)

Storks Male (12) 0.840 0.390 0.810 0.490 1.120 0.380 1.500

Storks Female (9) 0.720 0.600 0.470 0.370 0.800 0.230 2.000

Storks Total (21) 0.810 0.500 0.670 0.410 1.130 0.230 2.000

OREOCHROMIS (whole body section samples ppm wet weight)

NILOTICUS

Fish Ebb (8) 0.0055 0.0050 0.0025 0.0045 0.0070 0.0020 0.0100

Fish LM (3) 0.0030 0.0030 0.0000 0.0030 0.0030 0.0030 0.0030

Fish Murchison(7) 0.0077 0.0080 0.0022 0.0060 0.0090 0.0050 0.0110

Fish Total (17) 0.0059 0.0055 0.0027 0.0035 0.0078 0.0020 0.01 10

121 Table 5.2: Analysis of variance of total mercury concentrations in breast feathers of

African ﬁsh eagles (Haliaeetus vocifer) (n = 33) from Lake Mburo and Lake Victoria near Entebbe, Uganda (ppm dry weight).

122 Table 5.2. Analysis of variance of mercury concentrations in ﬁsh eagle feathers (n = 33) (ppm dry weight)

Variable Level 11 Mean (sd) F

Lak 18 .28 .14 Site 6 Mburo ( ) 17.79 .0002 Lake Victoria 15 .80 (.50)

Adult 15 .59 (.55) Age .84 .3671 Nestling 18 .45 (.30)

Female 13 .43 (.22) Gender .82 .3709 Male 20 .57 (.53)

Female 6 .38 .15 Gender (Adults) ( ) 1.45 .2505 Male 9 .73 (.68)

Female 7 .47 .28 Gender (Nestlings) ( ) .03 .8549 Male 11 .44 (.33)

Site - - 18.13 .0002 Multivariable A88 ' " 1 -3 1 .2620 ANOVA Gender - - 1 .29 .2657 Overall - - 6.91 .0012

123 Table 5.3: Results of analysis of variance of breast feather mercury concentrations between Marabou storks (Leptoptilos crumeniferus) (n = 21) and African ﬁsh eagles

(Haliaeetus vocifer) (n = 15) residing in the Entebbe/Kampala region of Uganda (ppm dry weight).

Table 5.4: Results of univariable analysis of variance of total mercury concentrations in breast feathers of marabou storks (Leptoptilos crumeniferus) (n = 21) from Kampala,

Uganda (ppm dry weight)

Table 5.5: Results of univariable analysis of variance of mercury concentrations in

Tilapia ﬁsh (Oreochromis niloticus) (n = 18) ﬁ'om three sites in Uganda (ppm wet weight).

124 Table 5.3. Results of analySis of variance of feather mercury concentrations between avian species from Lake Victoria (ppm dry weight)

Variable Level n Mean (sd) F p

Species Aﬁ'i can f1 5 h eag l cs 15 .80 ( .50 ) 0.0 .9510 Marabou storks 21 .81 (.49)

Species - - 5. 17 .0272 Multivariable ANOVA Gender ' " .30 . 5864 Overall - - 2. 74 .0743

Table 5.4. Results of analysis of variance of feather mercury concentrations in marabou storks (ppm dry weight)

Variable Level n Mean F p (8d) ” m " ” mwféh—aieﬂ F ’9 5 3T" 5 - Gender 0 O .9549 Male 12 80

Table 5.5. Results of analysis of variance of total body mercury concentrations in Oreocltromis uiloa'cus fish (n = 18) from three sites in Uganda (ppm wet weight)

Variable Level n Mean (sd) F p

Lake Mburo 3 .0030

Lake Victoria 8 .0055 Site 5.06 .0209

Murchison 7 .0077 Fans

125 Figure 5.1: Total mercury in breast feathers of adult and nestling African ﬁsh eagles

(Haliaeetus vocifer) from Lake Mburo (n= 18) and Lake Victoria at Entebbe (n= 15),

Uganda (ppm dry weight).

126 Figure 5.1. Total mercury in breast feathers of adult and nestling African fish eagles (Haliaeetus vocifer) (n = 33) (ppm dry weight)

2.5 i

2 1 1. .. —.— Lake Victoria! a ; 5. - {1 - Lake Mburo 1 E 1.5 . ‘ a o E [-1 1 4 0&9\.)\

0.5 « 9/%’/\ 5‘0" \ Cl.

'a \19 o ‘ I. ‘u D- ", D d E El .' D U DO '3 Clo-:1 . . . o . o 5 10 15 20

Individual Eagle Sample Number

127 Chapter 6

Nest Site Habitat Characterization of African Fish Eagles (Haliaeetus

vocifer) from three sites in Uganda

128 Abstract

The objective of this study was to document selected characteristics of the nest site habitat of African ﬁsh eagles (Haliaeetus vocifer) in Uganda. Nest tree species, nest tree height, nest height within the tree and nest tree diameter at breast height were recorded. Crown class, canopy cover, distance to water, whether the nest tree was living, decaying or dead and evidence of disturbance were recorded for each nest tree. Nest site habitat characteristics of breeding areas of African ﬁsh eagles (Haliaeetus vocifer) were examined at Lake Mburo (n =13), Murchison Falls National Park (n = 12) and Lake

Victoria at Entebbe (n = 19). The most common tree species utilized for nesting were

Acacia sieberiana (83%) at Lake Mburo; Balanites aegyptiaca (50%) and Tamarindus indica (33%) at Murchison Falls and a mixture of species including Chlorophora excelsa

'at Entebbe (26%). Average nesting tree height was 11.9 i 3.6, 17.7 i 6.7 and 29.6 i 7.0 meters for Lake Mburo, Murchison Falls and Entebbe respectively. Average nest height was 9.7 i 3.9, 13.1 :t 5.5, and 22.9 i 5.8 meters for each site as above respectively. The average diameter at breast height of trees was 65 :1: 14, 91 i 37, and 145 i 67 cm for each site as above respectively. Fish eagles preferred nesting in living trees. Most breeding areas contained significant alteration of natural vegetation features by animals or man. Characterization of nest site habitat may assist in fiiture conservation planning for the African fish eagle in Uganda. Nest site characterization may also provide preliminary data necessary for development of the fish eagle as a biomonitor of environmental change.

129 Introduction

The African ﬁsh eagle (Haliaeetus vocifer) has been called the “the voice of

Africa” (Brown 1980). It has also been described as one of the best-known raptors throughout sub-Saharan Aﬁica (Brown 1980). Despite this, much of the knowledge of

African ﬁsh eagles is through observational records alone. Through observation, Brown

(1980) characterized fish eagle breeding areas as usually having one to three nests, generally in tall trees and felt “ ..... quite certain that the fish eagle selects its nest tree, if there is a choice”. The objective of this study was to measure selected nest site characteristics of African fish eagle nests at Lake Mburo, Murchison Falls and Lake

Victoria near Entebbe in Uganda, East Africa. Data was collected as part of research determining concentrations of selected environmental pollutants in the tissues of ﬁsh eagles at the study sites. A second project objective was to assess the potential of ﬁsh eagles to be utilized as a biomonitor species of environmental change. Nest site habitat characterization may aid in this assessment by providing a quantitative description of selected nest site habitat requirements.

Materials and Methods

Selected characteristics of the breeding area of Aﬁ'ican ﬁsh eagles (Haliaeetus vocifer) were examined at Lake Mburo (0°. 39' S, 30°. 57' E) (n =13), Murchison Falls

National Park (2°. 15' N, 31°. 38' E) (n = 12) and Lake Victoria near Entebbe (0°. 04' N,

32°. 26' E) (n = 19). A breeding area was deﬁned as approximately 100 square foot of land and water centered on a tree with a nest in it. To qualify, the nest had to have evidence of recent use such as eagles present in the tree, visible urates around the tree

130 base, nesting activity and presence of prey remains. Lake Mburo is a 6 km long freshwater lake situated in south western Uganda within a 256 sq. km national park.

Areas of Entebbe studied included the Lake Victoria coastline from Nfo Island (0°. 00'N,

32°. 26'E) to Kisubi Bay (0°. 05'N, 32°. 35'E). Areas of Murchison Falls National Park sampled were the Nile river from the falls to Wanseko at the Lake Albert delta. Nest sites were sampled between December 2002 and January 2003.

Tree heights were determined using the percentage scale of a clinometer (Suunto

Clinometer PMS SPC, Carlsbad, CA). A ﬁberglass tape (Keson model OTR-l8M-15,

Forestry Suppliers, Inc. Jackson, MO) was used to measure 50 - 100 ft from the base of the nest tree. If vegetation was impenetrable, a rangeﬁnder (TLR 75 Ranging

Rangefinder, East Bloomfield, NY) was used for the same purpose. Tree height was calculated fi'om the formula: (percentage to top of tree + percentage to bottom of tree) x distance from tree (ft) = Height of Tree (ft). Nest height was calculated by the same method except the measurement was taken at the bottom of the nest. All results were

converted to meters. Diameter of the tree at breast height (DBH) was measured with a

DBH tape (Forestry Suppliers, Inc. Jackson, M0) at 1.37 meters (41/2 ft) above ground on

the uphill side of the tree. Distance to water was deﬁned as the nearest straight line from

the nest tree to the shore measured with a ﬁberglass tape, or a rangeﬁnder, or if the

distance was great, by visual estimation. Elevation and site location were recorded using

a global positioning system (Eagle Expedition, Eagle Electronics, Catoosa, OK). Canopy

cover was determined by use of a spherical crown densiometer (Spherical Densiometer

Model C, Forest Densiometers, Bartlesville, OK). Readings were taken ten paces from

the tree facing north, south, east and west. Vegetation had to be 7 m or more above the

131 ground to be included. The readings were averaged then multiplied by 1.04. This ﬁgure was then deducted to give the total area of the canopy occupied. Tree species were

identiﬁed from the keys found in Eggeling (1951). Tree condition was assessed visually

and recorded in the following categories living, 25% decay, 50% decay, 75% decay or

dead. Crown class was deﬁned as the nest tree height relative to the trees in a 100 ft

radius immediately surrounding the nest tree. Trees were classiﬁed as dominant if the tree was taller than those immediately surrounding it, co-dominant if the tree canopy was the

same height as the trees immediately surrounding it, intermediate if the nest tree canopy

was lower than the trees immediately surrounding it and suppressed if the nest tree was

completely dominated by the trees surrounding it. Age class of the nesting tree was

classiﬁed as even if all of the trees within the breeding area were approximately the same

size (signifying one canopy layer) or uneven, meaning the trees were of differing ages

and sizes (signifying multiple canopy layers). Evidence of disturbance was classiﬁed as

human or animal disturbance in an area of 100 ft radius from the base of the nesting tree.

Descriptive statistics were done using Excel (Microsoft Excel, Microsoft Ofﬁce

2000 Professional. Microsoft Corporation, Redmond, WA). Direct correlation was used

to assess the correlation between two variables.

Results

Measurements of nest tree height, nest height, tree height above nest, DBH,

elevation and canopy cover for African ﬁsh eagle nests at Lake Mburo, Murchison Falls

and Lake Victoria near Entebbe are presented in table 6.1.

132 The most common tree species utilized for nesting were Acacia sieberiana (83%) at Lake Mburo; Balanites aegyptiaca (50%) and T amarindus indica (3 3%) at Murchison

Falls and Chlorophora excelsa at Entebbe (26%). Other trees utilized for nesting were

Euphorbia candelabra and two unidentiﬁed hardwood species at Lake Mburo; A. sieberiana at Murchison Falls and Antiarus aﬁ'icana moraceae, Newtonia buchanii,

Antiarus toxocara, Canarium frankfurta, an unidentiﬁed Eucalyptus species and an unidentiﬁed Ficus species at Lake Victoria near Entebbe.

Nest trees were highest and had a greater diameter at breast height at Lake

Victoria near Entebbe. Lake Mburo had the smallest and thinnest nesting trees.

Crown class was dominant on 57%, 66% and 57% of the trees for Lake Mburo,

Murchison Falls and Entebbe respectively. The remainder were classiﬁed as co-dominant

except one tree at Entebbe that was classiﬁed as intermediate. No nesting trees were

suppressed. The nesting areas were comprised of trees of uneven age classes at 100%,

25% and 84% of nest sites examined at Lake Mburo, Murchison Falls and Entebbe

respectively.

At Lake Mburo and Murchison Falls 92% of nesting trees were living and at

Entebbe 74% were classiﬁed as living. One tree at Lake Mburo was dead while 4 trees

(21%) at Entebbe showed 25% decay and one tree 75% decay.

133 Evidence of human or animal disturbance was present at within the breeding area of all the nest sites. Human disturbance within the breeding area was seen at 89% of the

Entebbe nest sites. Animal disturbance within breeding areas at Lake Mburo was largely due to hippopotamus and buﬁ‘alo trails. Hippopotamus and buffalo trails were also numerous at Murchison Falls with the addition that many nests were located within crocodile breeding areas.

Discussion

Acacia sieberiana at Lake Mburo and Balanites aegyptiaca and Tamarindus indica at Murchison Falls are numerically predominant tree types in the landscape at these sites. It can be inferred that the diversity of tree types chosen for nesting at Entebbe relects anthropogenic change to the landscape and ﬁsh eagle adaptation to these changes.

Trees at Entebbe were generally taller than trees at the other two sites, but not as numerous. Eagle nests appeared to be more numerous at Entebbe where there were

aggregations of tall trees (such as in the botanical gardens and animal breeding centre)

and/or where there was restricted human activity (Nfo Island and smaller islands with

difﬁcult human access). The fewer trees available at Entebbe may have meant less

available choice for nest site placement and thus adaptation to less than suitable habitat.

Visual observation suggested that territories at Entebbe stretched a greater distance from

shore than at the other sites. This greater territorial depth may have been associated with

use of taller trees or differences in prey species density between sites. Visual observations

(not scientiﬁc counts with methodology) suggest greater density and smaller territories of

ﬁsh eagles at Lake Mburo than the other two sites. Determination of territory size may be

134 based on available resources, including suitable habitat, and this assertion may in part explain differing eagle densities between the study sites.

Data on the distance to water could be misleading as the distribution is highly skewed to the right suggesting a few nests were located at some distance from water with the majority closer. There was a weak positive correlation (r2 = 0.18) between tree height and distance from water suggesting that higher trees maybe required or sought out the farther the nest was from the shore. This would be necessary for prey visualization.

Uneven age class may signify anthropogenic disturbance precipitating variable vegetation growth rates or it could indicate the predominance of one or a few tree species. The high percentage of even age classes seen at Murchison Falls may be related to the topographical features of the area or the differences between nest sites located on a river with a strong current versus lake-based, nest sites. Prey density could also affect nest site habitat preference.

The data presented can best be utilized to draw inferences into what may constitute suitable nesting site habitat for ﬁsh eagles. A more comprehensive statistical analysis, beyond the resources of this project is required to place many of the measurements into perspective. For example, the data would suggest ﬁsh eagles prefer to nest in living trees however this cannot be proVed as the number of living: dead trees within each study site was not determined. Examination of the data in relation to the prevalence of the characteristic within a study site is also required (eg. is there a predominance of even or uneven age classes of trees in general at Lake Mburo). The data

135 is therefore presented without analysis as baseline information that requires further, more intensive and wider ranging data collection before fithher conclusions can be made regarding fish eagle nest site habitat. However, in regards to nest tree species, nest tree height, nest height and canopy cover the data from each study site shows marked variation. One conclusion that may be drawn from this data is that fish eagles are an adaptable species able to nest under a fairly broad range of habitat parameters.

Further characterization of nest site habitat, utilizing this data may assist in future conservation planning for the African ﬁsh eagle in Uganda. Understanding, characterizing and quantifying the norm may assist in understanding, quantifying and interpreting deviations from the norm. In this way nest site characterization may also provide preliminary data necessary for development of the ﬁsh eagle as a biomonitor of environmental change.

136 References

Brown L. 1980 The African Fish Eagle. Bailey Bros. & Swinfen Ltd. Folkestone. England.

Eggeling WJ. 1951. The Indigenous Trees of the Uganda Protectorate. The Government Printer, Entebbe, Uganda.

137 Table 6.1: Nest site characteristics for Aﬁican ﬁsh eagle (Haliaeetus vocifer) nests at

Lake Mburo (n = 13), Murchison Falls (11 = 12) and Lake Victoria at Entebbe (n = 19),

Uganda.

‘ LM = Lake Mburo b MP = Murchison Falls National Park c LV = Lake Victoria at Entebbe

138 Table 6.1. Nest site characteristics for African ﬁsh eagle

LM (n =13)‘ MF(n=12)r LV (n =19)c TREE HEIGHT (M)

Mean :1: SD. ll.9i3.6, l7.7i6.7 29.6: 7.0

Median (range) 11.9 (91-189) 17.1 (12.8-30.2 29.9 (25-46.4) NEST HEIGHT (M) Mean i SD. 97:39 131:55 229:58

Median (range) 8.8 (64-165) 12.2 (8.8-24.4) 23.8 (17.7-34.8) TREE HEIGHT ABOVE NEST (M) Mean :t S.D. 22:15 46:27 64:15 Median (range) 2.4 (0.9.4.9) 4.6 (2.4-9.8) 5.8 (40207) ELEVATION (M) Mean :l: S.D. 15941144 959185 1448151 Median (range) 1582 (1570- 936 (922- 1456(1450- 1937) 1150) 1512) DBH (CM)

Mean i S.D. 65: 14 91 :37 145 :67 Median (range) 67 (52-87) 86 (60-166) 118 (94-287) DISTANCE TO WATER (M) Mean i S.D. 76.8 i 166 48 :1: 50.3 294:1: 355 Median (range) 12.2 (1.8-500) 36.6 (20-176) 73.4 (9.1-1000) CANOPY COVER ('/o) Mean :l: SD. 37:32 61:1:22 47130

Median (range) 42 (8-95) 69 (55-88) 43 (21-90)

139 Chapter 7

Assessing the potential of African Fish Eagles (Haliaeetus vocifer) and

Marabou storks (Leptoptilos crumeniferus) as biomonitors of

environmental change.

140 Abstract

The purpose of this study is to assess the suitability of using African ﬁsh eagles

(Haliaeetus vocifer) and marabou storks (Leptoptilos crumeniferus) as biomonitors of environmental change in Uganda. A study was designed to evaluate concentrations of persistent organic pollutants and mercury in African ﬁsh eagles and marabou storks in

Uganda. A series of criteria necessary for the successful development of a biomonitor species was established ﬁ'om the work of others and our own observations. With the results from this study and the existing body of literature, the suitability of marabou storks and ﬁsh eagles as biomonitor species was assessed against these criteria. African

ﬁsh eagles and Marabou storks met most of the criteria of a suitable biomonitor species.

However, both species failed to meet critical criteria. Chiefly, for fish eagles the reproductive cycle needs to be defined more clearly at the local level as does local variation in diet. Marabou storks require development of suitable capture methods for the adult portion of the population. Biomonitoring programs using marabou storks need to consider the wide ranging nature of the species diet, and its migratory behavior in its tropical range.

Introduction

The use of mammals, birds and reptiles as monitors of the potential effects of chemicals in the environment is well documented (Burger & Peakall 1995; Jeﬁ'ee et al

2001). For a species to be a suitable biomonitor certain conditions must be met. These include characteristics that are deﬁned at the class, order, genus and species levels for the animal being assessed. A characteristic at the class level may be that all birds molt

141 feathers and this uniquely avian tissue may be a useful indicator of whole body concentrations of certain heavy metals. A characteristic at the genus level might be the predominantly piscivorous diet and large body mass of members of the Haliaeetus genus.

Documentation of the endpoint that is to be quantiﬁed for the chemical being examined, and the species in which it is to be examined are also important. An example would be eggshell thinning and its relationship to DDE concentrations in the parent birds.

Establishing the relationship between an endpoint indicator effect and its hypothesized cause often requires laboratory toxicity testing on the species or a similar species.

Selecting a tissue type in the correct quantity at a suitable body site, and establishing a correlation between tissue chemical concentrations and concentrations known to cause an endpoint effect are also critical. Finally, to utilize a species as a biomonitor, knowledge of ecological structure and function, and form and cycling of the chemical within the environment being examined are required. To meet these criteria, research teams with multidisciplinary expertise may be required. Apart ﬁom directly measuring residues of chemicals of concern in an animal species, information can be gained by use of other monitoring methods, such as stable radio-isotopic analysis, which can document changes in energy or mass flows through ecological communities (Harding & Stevens 2001).

Chemical effects on an environment could be one of many reasons for changes in energy or mass ﬂows that can be documented by radio-isotopic studies. Water quality analysis can help determine how a chemical may behave under different levels of anthropogenic alteration to aquatic ecosystems.

142 We sampled ﬁfteen adult and eighteen nestling African ﬁsh eagles (Haliaeetus

vocifer), twenty nestling marabou storks (Leptoptilos crumeniferus) and eleven tilapia

ﬁsh (Oreochromes niloticus) from Lake Mburo and the Kampala/Entebbe region of

Uganda for a range of persistent organic pollutants and mercury. The objectives of the research were to document baseline concentrations of these chemicals in marabou storks

and African ﬁsh eagles, as well as collect other baseline data, such as plasma chemistry values, nest site habitat characteristics and body measurements. A ﬁrrther objective was

to make an initial assessment of the utility of African ﬁsh eagles and marabou storks as biomonitors of environmental health. We wish to report our conclusions in relation to this objective.

Materials and Methods

African ﬁsh eagles were sampled at Lake Mburo, a 6 km long freshwater lake in

South Western Uganda (0". 39' S, 30°. 57' E) situated in a 256 km2 national park, and on

Lake Victoria at Entebbe, from Nfo Island (0°. 00' N, 32°. 26' E) to Kisubi Bay (0°. 05' N,

32°. 35' E). Fish eagles were sampled at Lake Mburo in July, August and December 2002

and at Lake Victoria in August 2002 and January 2003. Thirty-three eagles were

sampled: ten adults and eight nestlings from Lake Mburo and ﬁve adults and ten

nestlings from Lake Victoria. Twenty marabou stork nestlings were sampled from a

colony located along Nile Avenue, in central Kampala, Uganda (0°. 32' N, 32°. 58' E).

Marabou stork nestlings were sampled from twelve nests, in six trees (Tabebuia pentaphylla), in one colony, on one day from 0700 till 1800, in January 2003.

143 Oreochromis niloticus were sampled at Lake Mburo and Lake Victoria at Entebbe in the same locations and times that ﬁsh eagles were sampled.

Adult ﬁsh eagles were captured on water using a ﬁsh “snare vest” technique.

Oreochromis niloticus were fitted with eight to twelve 5 - 6 cm diameter fishing line snares (loops). The free ends of the snares penetrated the body of the fish and were then tied on themselves and the excess line cut. The line was attached to a hand held wooden reel in a boat. The snared fish was then placed in water. Once captured, the eagle was taken to shore and placed in dorsal recumbency. Ten ml of blood were collected from the brachialis vein via a 21 or 23 gauge x 1.9 cm (3/4 inch) butterfly catheter (Surfio Winged

Inﬁrsion Set, Elkton, MD) connected to a 10 ml syringe (Luer Lok Tip Syringe, Becton

Dickinson and Company, Rutherford, NJ). The blood was immediately transferred to a 10 ml lithium heparin vacutainer. An additional 4 ml of blood was drawn and placed in a 5 ml EDTA vacutainer (Becton Dickinson, Franklin Lakes, NJ). A drop of whole blood was placed on a commercially prepared paper sample card for molecular sex determination (Avian Biotech International, Tallahassee, FL). Five whole breast feathers were then hand plucked for determination of total mercury concentrations. Body measurements and weights were recorded, blood smears made and the bird banded.

Eagles were released from land at the closest point possible to the capture location.

Average time ﬁ'om capture to release was 34 minutes (range 20-45 minutes).

African ﬁsh eagle and marabou stork nestlings were retrieved for sampling from

the nest using professional tree climbing techniques (USDA Forest Service, 1996) with

144 modiﬁcations for tropical tree species and environmental conditions. The main method of tree ascent was using tree climbers (Klein Tools, Chicago, ILL). Eagle and stork nestlings were placed singly into a 40 cm diameter nylon bag (The Taku Tailor, Juneau, Alaska,

USA) and lowered to the ground where sampling occurred. Sampling of nestlings was as described for adults with the exception that the volume of blood collected was between 4 to 14 ml depending on body weight. No more than one percent of body weight was collected and often considerably less. Venipuncture site on stork nestlings was the medial metatarsal vein. Average time from removal of the nestling until return was 14.5 minutes

(range 7-22) for marabou storks and 32 minutes (range 10-60) for nestling ﬁsh eagles.

Oreochrornis niloticus samples were obtained fi'om local fisherman as close as possible to where eagles were sampled. Fish less than 300-400g and 300-450mm in length were selected. Fish were weighed and their body length recorded. A 100g whole body section sample of fish was dissected just cranial to the dorsal fin and weighed. The sample included skin, muscle, bone and viscera.

There are perhaps no species that would meet all the criteria of an ideal biomonitor. To test the suitability of developing the Aﬁ'ican ﬁsh eagle and marabou stork as biomonitors of environmental pollution the following criteria were established from the work of others (Gragnaniello et al 2001; Burger and Peakall, 1995, State of the Lakes

Ecosystem Conference, SOLEC 1998; International Joint Commission. 1985) and our own observations. We believe the list constitutes the ideal for a biomonitoring program utilizing avian species. African ﬁsh eagles and marabou storks were subjectively assessed on their ability to meet the following requirements:

145 . The population of the species must be large enough such that sampling will not

adversely affect the population.

The species should be nonmigratory, at least for the part of their life cycle when

sampling occurs, so as tissue concentrations of chemicals are an indication of local

environmental contamination.

The biology of the species has been characterized and changes can be monitored.

The species should be large enough so that adequate samples can be obtained to

meet analysis requirements.

Size, age and sex differences within the species are capable of being documented.

Size, age and sex variation in bioaccumulation of the chemical within the species

can be documented.

The diet of the species can be determined for the environment under examination.

Species foraging range must be known, if local point source determinations are

required.

The diet of the species must be relatively consistent within and between

environments under study.

10. The species reproductive cycle must be known, and reproductive success and

productivity able to be determined quantitatively. ll. Exposure routes for the chemical in the species concerned need to be documented.

12. An endpoint effect that is to be quantiﬁed for the chemical must be demonstrated in

the species examined, or a similar species under experimental laboratory conditions.

146 l3. Species must have a specimen that can be sampled, in the correct quantity at a

suitable body site, in which there is a correlation between concentrations found and

concentrations known to cause a speciﬁc endpoint.

14. The specimen chosen for sampling must be able to be stored from the time of

sampling until analysis in a manner that will not affect the analysis or results.

15. The species can be sampled cost effectively and with relative ease, in the

environment under examination.

16. If bioaccumulation of the chemical is to be studied, the species must occupy an

upper trophic level in the food chain.

17. The species should be able to tolerate a range of concentrations of the chemical

under examination.

18. Public and regulatory acceptance of the species as a biomonitor and the sampling

methods utilized should be established.

19. Knowledge of the chemicals to be examined, their localized usage, their activity and

reactions in the environment are required.

20. The species should have a long life span.

21. The species should be monitored over a number of seasons or biological cycles.

Results

African ﬁsh eagles partially meet some of the criteria used to characterize an ideal biomonitor species. In summary, the population is large and secure at the study sites sampled, the species is nonmigratory as adults, the diet is known, the species is long lived, sex and size differences can be accounted for, sufﬁcient volumes of blood and

147 feathers are available for analysis, the species occupies a high trophic level in the food chain and laboratory toxicity and field studies are available on similar species. In addition sampling could be cost effective once capital equipment is purchased and appropriate and willing collaborative partnerships are established. Criteria that have not been met are the biology of the African fish eagle is not well documented, nor is the reproductive cycle. In addition the diet of the fish eagle may not be consistent between locations.

Similarly to ﬁsh eagles, marabou storks meet some of the criteria of a suitable biomonitor. Marabou storks have large, relatively secure populations in Uganda, their biology and reproductive cycle has been recorded, sufﬁcient sample volumes can be obtained, nestlings are relatively easy to sample, a portion of the population is nonmigratory (nestlings) and the species is long lived in the wild. However, adult marabous are mainly migratory, adult birds are very difficult to sample and the diet may not be consistent within and between sites.

Discussion

Fish eagle biology and biomonitoring

African fish eagles meet some of the criteria used to characterize an ideal biomonitor species. Afiican fish eagles are tertiary avian predators in lake-based ecosystems and thus occupy a high trophic level in the foodchain (Stewart 1997). Local populations of fish eagles at both study sites appear adequate for sampling purposes.

Population counts at both sites have been conducted approximately twice yearly (The

East Aﬁica Natural History Society, PO. Box 27034, Kampala, Uganda) and suggest that

148 the population is stable. Fish eagles are thought to be nonmigratory as nestlings and as mature adults with an established territory (Brown 1980). Therefore, adults with established territories and nestlings should reﬂect local concentrations of contamination.

One study related observations of productivity in fish eagles on a lake polluted with heavy metals and pesticides and made the observation that the fish eagle is considered to be a useful indicator of pollution in aquatic ecosystems (Mundy & Couto 2000). One of the key factors in making this assessment was the sedentary nature of adult fish eagles.

The biology of the ﬁsh eagle at the two sites examined has not been scientiﬁcally

studied. Scientiﬁc studies at other sites are minimal, scattered and not comprehensive.

Much of the knowledge of ﬁsh eagle biology has been through records of observations,

rather than planned studies. Without a sound knowledge of what is normal it is difﬁcult to recognize and assess variation that may be associated with environmental

contamination.

Fish eagle diet and biomonitoring

Adult ﬁsh eagles in this study weighed 2.2-3.6 kg. Nestling ﬁsh eagles could be safely

sampled at a body weight of 800 grams or above. The lower nestling weight gave enough

plasma (3 ml) to complete all tests required. Avian sex in non-ratite species can easily be

determined with almost 100% accuracy by examination of chromosomal DNA (Grifﬁths

et a1 1998). As DNA is present in all cells, commercial tests are available using blood and

feather samples. Once adult plumage has been established at approximately 5 years, it is

impossible to age ﬁsh eagles (Brown & Cade 1972). Although most studies with feathers

have failed to ﬁnd a relationship between mercury and bird age, Gochfeld et a1 (1996)

149 found that mercury concentrations in feathers of laughing gulls (Lams atricilla) decreased significantly with age. This highlights the importance of banding eagles so band returns may give some indication of longevity in the wild, which is unknown but calculated to be 19.8 years (Brown and Cade 1972). Thus fish eagles can be sexed, can be aged as nestlings, juveniles or adults and are large enough to obtain sufficient sample volumes for most types of analysis.

African ﬁsh eagles feed on a variety of ﬁsh with cichlid species such as tilapia

(Oreochromis niloticus) that occupy the littoral zone comprising 80% of the ﬁsh taken

(Stewart et a1. 1997). Oreochromis niloticus were chosen for sampling as it appeared to be the fish eagles main prey item at Lake Mburo and Entebbe. Fish eagles are exposed to both mercury and persistent organic pollutants by dietary exposure, with bioaccummulation and biomagnification increasing contaminant concentration with increasing food chain trophic level. However our observations and anecdotal reports suggest that fish eagle diet may vary between sampling locations and even between individual birds. Our observations of prey remains at nests in Murchison Falls indicate tiger fish (Hydrocynus forsaklii), a carnivorous species often occupying surface waters forms a large percentage of the fish eagle diet in this area. In other areas, it is suggested some fish eagles are largely orniphagous (Brown, 1980). If carnivorous avian and fish species were the main prey items, this would mean another trophic level added to the food chain that may result in greater exposure of fish eagles to harmful chemicals through bioaccumulation. Such variation in diet between sites may complicate comparisons of chemical effects on indicator species between sites.

150 Fish eagle reproduction and biomonitoring

Details of the reproductive biology of the ﬁsh eagle have been reported and observed by

Brown (1980) but not scientiﬁcally evaluated. Sumba (1986) examined breeding seasonality in ﬁsh eagles at Queen Elizabeth National Park. Sumba concluded that breeding in Queen Elizabeth National Park was non seasonal while Brown (1980) stated

“...in some areas ﬁsh eagles breed at any time of the year; in others they are more strongly seasonal, though even the breeding season tends to be rather elastic.” Our observations tend to agree with this statement. The breeding cycle of ﬁsh eagles at both sites in this study has not been characterized nor has reproductive success and productivity been measured over a number of years. Our observations at Lake Mburo suggest breeding at this site may be seasonal. No chicks were observed in nests during

December or January. Chicks ranging from approximately one week old to almost

fledged were found in nests during late June, July and August. Breeding at Lake Victoria near Entebbe may be less seasonal. Chicks ranging from approximately one week old to almost fledged were found in nests in July and August while chicks fiom one week to approximately 4 weeks old were found in nests in January and February. Lack of a defined seasonality to the breeding cycle indicates that any monitoring program ideally needs to have the logistical capacity to sample at a numerically low level over extended periods of time. A monitoring program needs to establish breeding patterns of the indicator species at the specific study site, rather than relying on data extrapolated from other areas. Population reproductive measures such as productivity (the number of young per occupied nest for each breeding area) and success (percent of occupied breeding areas

151 successfully fledging at least one young), as used in studies on bald eagles (Bowerman,

1993) provide useful information only if they are conducted as multiyear analyses.

Laboratory and field toxicity studies: mercury

Altered nesting behavior and reproductive failure have been documented in wild common

loons (Gavia immer) with elevated mercury concentrations (Barr 1986). Laboratory

feeding studies on Goshawks (Accipiter gentiles) and red-tailed hawks (Buteo jamaicensis) with mercury have shown acute death, neurotoxicity and altered nesting

behavior (Borg et al 1970; Fimreite 1971). Scheuhammer and Bond (1991) suggest that

mercury feather concentrations greater than 20 ppm may be associated with toxic effects.

Eisler (1987) states that concentrations above 5ppm fresh weight in sensitive avian

species are thought to be associated with adverse affects. However, toxic effects of

mercury may vary among bird species (Gouter et a1 1998).

Feathers as a sample for biomonitoring mercury

Feathers are an excellent, non- invasive tissue to sample for mercury analysis. Feathers

contain higher concentrations of mercury than in other body organs (Westermark et al

1975) with approximately 70% of total body mercury found in feathers. (Honda et al

1986). A positive correlation has also been shown between mercury content in feathers

and internal tissues (Thompson et a1 1991). Mercury in feathers is strongly bonded to

keratin disulphide bonds (Crewther 1965) and does not easily degrade therefore archived

feathers can be used for analysis. Mercury concentrations are spread evenly throughout

the feather, indicating endogenous incorporation methods, rather than atmospheric

152 deposition (Hahn et al 1993). Feathers are a major excretory pathway for body mercury during molt. Differing mercury concentrations are found depending on the order the sampled feather has in the molting sequence. This has not been consistently shown to occur in all individuals within a population. Body contourfeathers are the feather type that shows the least variation in mercury content (Fumess et a1 1986). The methodology of feather sampling used in this study was based on these ﬁndings.

Persistent organic pollutants, ﬁsh eagles and biomonitoring

The effects of the pollutants tested for in this study have been well documented in laboratory toxicity studies and ﬁeld reports on analogous raptor species, including the bald eagle (Haliaeetus Ieucocephalus) (Bowerman 1993; Lincer 1975). Eggshell thinning has been used as an indicator of toxic effects of DDE in avian species with 18 to 20% being related to declines in populations (Blus et al 1996). Egg concentrations of DDT have been positively correlated with blood concentrations. Sampling of blood can therefore give an indirect indication of egg concentrations of the contaminant (Henny and

Meeker 1981). Determination of the lipid content of the plasma is important in that a

study by Elliot et al (1998) found a positive correlation between mean plasma lipid in nestlings and productivity (Elliot et al 1998). Plasma concentrations of POPs need to be

analyzed in relation to population reproductive success and productivity to assess the

effect of the chemical. This was not possible within the limited temporal period of this

project.

153 Fish eagle sampling methodology and biomonitoring

Fish eagle nestlings are difﬁcult to sample. Nests are located in tree species that are challenging to climb and require professional tree climbing equipment. Nests are

fi'equently located on thin branches that require ingenuity and resourcefulness on the part of the climber to reach. Occasionally, other complications such as wasps, ants and the tree being too tall for the equipment were encountered. The majority of trees at both sites were accessible only by water. Climbing the nesting tree was the only accurate way to determine whether a nest contained chicks. However, the nests were generally very visible and within a reasonable distance of each other, thus facilitating logistics. The use of a fish snare vest may be an effective method to catch adult African fish eagles but the success rate is very site dependent due to a host of multi-factorial local site conditions.

Marabou storks as biomonitors

Marabou storks provide an interesting comparison to ﬁsh eagles when assessing the species potential as biomonitors. In contrast to the ﬁsh eagle, the biology and reproductive cycles of the marabou stork in the study area have been well-documented

(Pomeroy 1977, 1978). Marabou storks have a well-defined breeding season and are communal nesters. Multiple nests can occur in a single tree. Marabou storks have an extremely long fledging period, with first flights out of the nest occurring at 110-115 days (Hancock et al 1992). Marabou stork nestlings, when they are unable to fly, are relatively easy to sample. Nesting trees generally present less of an access challenge than

ﬁsh eagle nests. Marabou parents can vigorously defend their young but, like ﬁsh eagles, seem minimally affected by the sampling process. Once the location of a colony is

154 known, multiple nests in a single tree can facilitate rapid sampling of individuals. Nesting trees are often used for many years making estimations of population reproductive parameters easier to assess. Marabou storks ﬁrst breed at 6-7 years and may live up to

25 years. Marabou storks are large, conspicuous birds, with average adult weights recorded being 5.66 kg for females and 7.06 kg for males (Pomeroy 1977). All these factors make marabou storks an eminently suitable biomonitor species. However, marabous have adapted to human activities by adopting a scavenging lifestyle. The diet of the birds in Kampala probably includes almost anything organic, such as garbage, ﬁsh remains, abattoir reﬁrse and a large amount of vegetable matter (Brown, et al 1982).

However, during the period of nestling growth increased amounts of protein are taken in the form of ﬁsh, frogs and rodents. The cosmopolitan nature of the urban marabou’s diet makes assessing sources of dietary exposure and establishing bioaccumulation in the food chain difﬁcult. In addition, marabous undertake seasonal, mainly north south migrations in Uganda, coinciding with rainfall seasonality (Pomeroy 1977). Although marabou storks are present year round in Kampala, the majority of the population is migratory.

Leg bands to identify non-migratory birds would be difficult due to the bands being obscured by urate wastes coating the legs of this species. Other methods of permanent adult identification have proved difficult and challenging (Pomeroy 1975).

Concentrations of chemical residues in adult marabou stork tissues may not be wholly due to local exposure. This problem is partially overcome by sampling nestlings. In addition, adult marabou storks are very difﬁcult to capture. Cannon nets, drop nets

(baited for many months to allow habituation) and use of oral anesthetics may be the only methods to achieve multiple captures. All these methods have their disadvantages. Oral

155 anesthetics have been tried previously with success but also caused a number of mortalities (Pomeroy and Woodford 1977). Our study unsuccessfully attempted capture with drop nets and hand held nets. Interestingly, Nyangababo (2003) was able to capture greater than 30 birds by using a piece of meat or ﬁsh as bait then hand catch the bird. We tried and failed with this method at many sites as marabous are extremely cautious and quickly learn to differentiate between a human bypasser and one attempting to capture them. Marabou storks also have the disadvantage as a biomonitor that laboratory based avian heavy metal toxicity trials have mostly been conducted in raptors and domestic avian species, with no known laboratory studies on stork species. With rapid increases in motor vehicle numbers and usage and poor quality grades of fuels being used in Uganda

(Uganda Bureau of Statistics, 2002) the marabou stork could be utilized in well-designed and reported ﬁeld research to monitor pollutants such as lead, cadmium and copper. The marabou stork could also be used to establish a pathology database, as many injured birds have to be euthanased each year. A number of studies in other avian species have assessed the use of feathers as biomonitors of atmospheric pollutants that have a component derived from exogenous deposition (rather than dietary exposure alone), such as atmospheric lead (Dauwe et al 2002; Wren et a1 1994; Nyagababo 2003).

Conclusion

We conclude that African ﬁsh eagles could be valuable monitors of selected environmental pollutants if the following preconditions are met:

1. Any study should be a multiyear analysis that requires a minimum low-grade

constant surveillance component.

156 . Any study should initially address the issue of more completely characterizing the

biology of the ﬁsh eagle, with particular emphasis on reproduction.

. Preliminary research should address what contaminants are of most concern in

Uganda and can the ﬁsh eagle serve as an appropriate monitor speciﬁcally for these

pollutants.

. Research programs should be designed to collect as much peripheral information as

possible (e.g. parasite loads and species, serological exposure to various avian

pathogens) and establish linkages with other investigators that can utilize this

information. This not only increases the physiological database for the species but

maximizes the use of ﬁeld resources and project budgets.

. Adult capture techniques should be reﬁned so they work consistently at multiple

study sites.

. Prey item selectivity for ﬁsh eagles and marabou storks needs further

characterization by prey remain studies at multiple study sites.

157 7. Appropriate collaborative partnerships with individuals, institutions and regulatory

authorities within Uganda should be developed. Resource allocation to provide for

local personnel development and training should be established with the goal of

devolution of management responsibilities to local personnel.

One way to achieve some of these goals may be integration with other ﬁeld research

projects such that resources and personnel are shared.

Marabou storks could be developed as a biomonitor species with the precondition that development of any research program recognize the advantages and disadvantages

discussed above and that the research is designed accordingly.

An area of research complimentary to residue analysis would be examination of

temporal and spatial dietary shiﬁs in marabou storks or ﬁsh eagles utilizing stable radio-

isotopic analysis (Harding & Stevens 2001). The shifts could then be correlated with

changes in many parameters, such as population density, chemical use or habitat

degradation. Water quality analysis in aquatic biomonitoring programs would also be

useful. This would enhance knowledge gained from residue analysis alone.

158 References

Barr JF. 1986. Population dynamics of the common loon (Gavia immer) associated with mercury-contaminated waters in northwestern Ontario. Occasional Paper No. 56. Canadian Wildlife Service. Ottawa.

Blus LJ, Weimeyer SN & Henny CJ. 1996. Organochlorine pesticides. In A. Fairbrother, LN Locke & GN Hoff (Eds) Noninfectious diseases of wildlife. Iowa State University Press. Ames, Iowa. pp. 61-70.

Borg K, Eme K, Hanko E & Wanntrop H. 1970. Experimental secondary methyl mercury poisoning in the goshawk (Accipiter gentiles). Environ. Pollut. 1: 91-94.

Bowerman WW. 1993. Regulation of Bald Eagle (Haliaeetus leucocephalus) Productivity in the Great Lakes Basin: An Ecological and Toxicological Approach. PhD Theses. Michigan State University.

Brown L. 1980. The African Fish Eagle. Bailey Bros. & Swinfen Ltd., Folkestone, England.

Brown L H & Cade TJ. 1972. Age classes and population dynamics of the bateleur and Aﬁican ﬁsh eagle. Ostrich. 43: 1-16

Brown L H, Urban EK & Newman K. 1982. The Birds of Aﬁica. 1. Academic Press, New York.

Burger J & Peakall D. 1995. Methods to assess the effects of chemicals on aquatic and terrestrial wildlife, particularly birds and mammals. In: Methods to Assess the Effects of Chemicals on Ecosystems. RA Linthurst, P Bourdeau & RG Tardiff (Eds) John Wiley and Sons Ltd. Chichester, England. Pp. 291-306.

Crewther W G, Fraser RDB, Lennox FG & Lindley H. 1965. The chemistry of keratins. Advanced Protein Chem. 20: 191- 346.

159 Dauwe T, Bervoets L, Blust R & Eens M. 2002. Tissue levels of lead in experimentally exposed zebra ﬁnches (Taeniopygia guttata) with particular attention on the use of feathers as biomonitors. Arch. Environ. Contam. Toxicol. 42: 88-92.

Eisler R. 1987. Mercury hazards to ﬁsh, wildlife and invertebrates: a synoptic review. US Fish and Wildlife Service Biological Report 85 (1.10), 90pp

Elliot JE, Moul IE & Cheng KM. 1998. Variable reproductive success of bald eagles on the British Columbia coast. J. Wildl. Manag. 62: 518-529.

Fimreite, N., & L. Karstad. 1971. Effects of dietary methylmercury on red-tailed hawks. J. Wildl. Manage. 35: 293-300.

Fumess RW, Muirhead SJ & Woodbum DM. 1986. Using bird feathers to measure mercury in the environment: relationships between mercury content and moult. Mar. Pollut. Bull. 17: 27-30.

Gochfeld M, Belant JL, Shukla T, Benson T & Burger J. 1996. Heavy metals in laughing gulls: gender, age and tissue differences. Environ. Toxicol. & Chem. 15: 2275- 2283.

Goutner V, Fumess RW & Papaponstantinou K. 2000. Mercury in feathers of Audouin’s gull (Len-us audounii) chicks from northeastern Mediterranean colonies. Arch. Environ. Contam. Toxicol. 39: 200-204.

Gragniello S, Fulgione D, Milone M, Soppelsa O, Cacace P & Ferrara L. 2001. Sparrows as possible heavy-metal biomonitors of polluted environments. Bull. Environ. Contam. & Toxicol. 66: 719-726.

Griffiths R, Double MC, Orr K & Dawson RJ. 1998. A DNA test to sex most birds. Molecular Ecol. 7: 1071-1075.

160 Hahn E, Hahn K & Stoeppler M. 1993. Bird feathers as bioindicators in areas of the German environmental specimen bank- bioaccumulation of mercury in food chains and exogenous deposition of atmospheric pollution with lead and cadmium. Sc. Total Environ. 139/ 140: 259-270.

Hancock J A, Kushlan J A. & Kahl MP. 1992. Storks, Ibises and Spoonbills of the World. Academic Press Lim. San Diego, California, USA. Pp. 133-138.

Harding E K & Stevens E. 2001. Using stable isotopes to assess seasonal patterns of avian predation across a terrestrial-marine landscape. Oecologia. 129: 436-444.

Henny C J & DL Meeker. 1981. An evaluation of blood plasma for monitoring DDE in birds of prey. Environ. Pollut. (A) 25: 291-304.

Honda K, Nasu T & Tatsukawa R 1986. Seasonal changes in mercury accumulation in the black cared kite, Milvus migrans lineatus. Environ. Pollut. (A) 42: 325-334.

International Joint Commission. 1985 Report of the Aquatic Ecosystem Objectives Committee to the Great Lakes Scientiﬁc Advisory Board. Windsor, Ontario.

Jefree R A, Markich SJ & Twining JR. 2001. Element concentrations in the ﬂesh and osteoderms of estuarine crocodiles (Crocodylus porosus) from the Alligator river regions, northern Australia: biotic and geographic effects. Arch. Environ. Contam. Toxicol. 40: 236-245.

Lincer J. 1975. DDE induced eggshell thinning in the American kestrel; a comparison of the ﬁeld situation and the laboratory results. J. App. Ecol. 12:16-21.

Mundy P J & Couto J T. 2000. High productivity by ﬁsh eagles on a polluted dam near Harare. Ostrich. 71: ll-14.

161 Nyagababo JT. 2003. Trace metals in the contour feathers of marabou stork (Leptoptilos crumeniferus) from Kampala city and its surrounding areas. Bull. Environ. Contam. Toxicol. 70: 792-799.

Pomeroy DE. 1975. Marking marabou storks. Bird Banding. 46: 343 — 344..

Pomeroy DE. & Woodford MH. 1976. Drug immobilization of marabou storks. J. Wildl. Manage. 40: 177-179.

Pomeroy DE. 1977. The biology of marabou storks in Uganda. 1. Some characteristics of the species, and the population structure. Ardea. 65: 1-24.

Pomeroy DE. 1978. The biology of marabou storks in Uganda. II. Breeding biology and general review. Ardea. 66: 1-23.

Scheuhammer A M. & Bond D. 1991. Factors affecting the determination of total mercury in biological samples by continous-ﬂow cold vapor atomic absorption spectrophotometry. Biolog. Trace Elem. Res. 31: 119-129.

SOLEC. 1998. Selection of indicators. State of the Lakes Ecosystem Conference 1998. N. Stadler-Salt and P. Bertram (Eds) Hamilton, Ontario, Canada.

Stewart K M, Matthiesen DP, Leblanc L & West J. 1997. Prey diversity and selectivity of the African ﬁsh eagle: data from a roost in northern Kenya. Afric. J. Ecol. 35: 133- 145.

Sumba SJA. 1986. Breeding seasonality of the African ﬁsh eagle in Queen Elizabeth Park, Uganda. Aﬁic. J. Ecol. 24: 103-110.

Thompson DR , Hamer KC & Fumess RW. 1991. Mercury accumulation in great skuas Catharacta skua of known age and sex and its effects upon breeding and survival. J. App. Ecol. 28: 672-684.

Uganda Bureau of Statistics. 2002. Key Economic Indicators. 47"1 Issue. Ch 39. Uganda Bureau of Statistics, Entebbe, Uganda.

162 United States Department of Agriculture Forestry Service. 1996. National tree climbing ﬁeld guide, USDA Forest Service. Missoula Technology and Development Center, Missoula, Montana, 1996.

Westermark T, Odsjo T & Johnels AG. 1975. Mercury content of bird feathers before and after Swedish ban on alkyl mercury in agriculture. Ambio. 4: 87-92.

Wren CD, Nygard T & Steinnes E. 1994. Willow Ptarmigan (Lagopus lagopus) as a biomonitor of environmental metal levels in Norway. Environ. Pollut. 85: 291-295.

163 Chapter 8

Project Conclusions and Recommendations for Future Research

164 Conclusions

In the research project we demonstrated that:

0 There is no signiﬁcant difference in persistent organic pollutant and mercury

concentrations in tissues of African Fish Eagles (Haliaeetus vocifer) and tilapia

(Oreochromis niloticus) from Lake Victoria near Entebbe compared to those from

Lake Mburo.

We found a signiﬁcant (p s 0.05) difference between concentrations of mercury in African ﬁsh eagle feathers as well as whole body cross section samples of

Oreochromis niloticus collected from Lake Mburo and Lake Victoria near Entebbe. The mercury concentrations in African ﬁsh eagle feathers and whole body section samples of

Oreochromis niloticus ﬁ'om Lake Victoria were consistemly higher than those from Lake

Mburo. Feather mercury concentrations in ﬁsh eagles from Lake Victoria were still signiﬁcantly higher (p = 0.0012) than concentrations from Lake Mburo when multivariable analysis was performed to account for gender and age (nestlings or adults).

Further to this we found no significant (p s 0.05) difference in feather concentrations of mercury in marabou stork nestlings from Kampala and those from African fish eagles from Lake Victoria near Entebbe. We therefore find the alternative hypothesis proved (at the 0.05 level of significance) and reject the null hypothesis in relation to mercury concentrations. The higher mercury levels at Entebbe and Kampala are most likely the result of anthropogenic emissions, particularly the use of fossil fuels as well as the burning of garbage. Afiican fish eagle plasma, marabou stork plasma and whole body section samples from Oreochromis niloticus did not contain the following chemicals at

165 the 0.001 ppm limit of detection: aldrin, DDT, a-HCH, dieldrin, endrin, heptachlor and their metabolites, B-HCH, 2,4’-DDD, 4,4’-DDD, 2,4’-DDT, 4,4’-DDT, heptachlor epoxide and lindane and nonachlor. Total PCBs were not detected in African ﬁsh eagle plasma, marabou stork plasma or whole body section samples from Oreochromis niloticus at the 0.003 ppm limit of detection. Five adult eagles from Entebbe had 4,4'-

DDE detectable in plasma, one at 0005me one at 0.003ppm. two at 0.002ppm and one at 0.001ppm wet weight. Five Oreochromis niloticus samples from Entebbe contained

4,4'-DDE concentrations of 0.001, 0.001, 0.002, 0.003 and 0.003 ppm wet weight. At the

0.001 ppm limit of detection 4,4'-DDE was not detected in nestling marabou stork plasma samples fi'om Kampala. We therefore find the alternative hypothesis proved and reject the null hypothesis (at the 0.05 level of significance) in relation to 4,4'-DDE but accept the null hypothesis in relation to all other sampled pesticides and PCBs.

Important issues related to the hypothesis that were addressed by this research were:

0 Whether concentrations of persistent organic pollutants and mercury are likely to be

major contributing factors to the population dynamics of the Aﬁican ﬁsh eagle

(Haliaeetus vocifer) at Lake Victoria near Entebbe or at Lake Mburo.

Reproductive parameters were not quantified for the African fish eagle populations as this would require multiyear analysis. However the concentrations of mercury and 4,4'-DDE found in Afiican fish eagle feathers and plasma respectively are below concentrations known to cause toxic and reproductive change in a variety of avian

166 species under ﬁeld and laboratory conditions. Our observations of African ﬁsh eagle nest productivity, breeding success and population structure lend support to this statement.

However with an adult sample size of five birds fi'om Entebbe, the statement should be interpreted cautiously in relation to our results. To definitively investigate the relationship between chlorinated hydrocarbons and productivity in African fish eagles a larger adult sample size, continued nestling sampling and long-terrn population reproductive analysis is required. Also required is greater knowledge of the clinical and sub clinical effects of mercury on piscivorous avian raptorial species, such as the African fish eagle and

C iconiini, such as the marabou stork.

0 Whether African ﬁsh eagles (Haliaeetus vocifer) and marabou storks (Leptoptilos

crumeniferus) can be utilized as successﬁrl biomonitors of mercury and persistent

organic pollutants.

We found that African fish eagles and marabou storks met most of the criteria of a suitable biomonitor species as stated in this thesis. However, both species failed to meet critical criteria: the biology of the African fish eagle is poorly understood, the reproductive cycle has not been characterized, the reproductive cycle and diet may show site variation, nestlings are difficult to sample and the success of methods utilized to catch adults appears dependant on multi-factorial local site conditions. The marabou stork is migratory, has a varied diet often based on scavenging off human waste and the adult birds are extremely difficult to capture for sampling. With development and research most of these impediments to meeting the stated criteria could be met. We

167 conclude that both species show promise as biomonitors of chemical and heavy metal contaminants but require more development. In the case of the fish eagle, firrther research investigating species biology and reproduction is required. For the marabou stork, research projects require a design and purpose that can achieve project objectives within the limitations of species-specific characteristics (e.g. the species is migratory, varied diet).

Recommendations

Relative political and economic stability in Uganda for the last 16 years has led to increased foreign aid and investment, industrialization, urbanization and centralization of government. The fact the human population of Kampala increased fi'om 774,241 in 1991 to a preliminary figure of 1,208,544 in 2001 and use of diesel firel (inferior grades and quality) in Uganda increased fi'om 125,621 in 1997 to 207,183 cubic metres in 2001

(Uganda Bureau of Statistics, 2002) lend support to these assertions. There are a number of regulatory authorities charged with the management of Uganda’s environment. To be proactive in their respective environmental mandates these authorities require baseline data to establish and monitor trends in environmental indices. The use of mammals, birds and reptiles as biomonitors can form part of this process. With this in mind the following recommendations are made in relation to this research project:

In relation to the concentrmions of mercury and persistent organic pollutants found in

African ﬁsh eagles, marabou storks and Oreochromis niloticus ﬁsh, no remedial management actions are advised.

168 In relation to the development and use of African eagles and marabou storks as biomonitors the following is advised:

1. Develop a multiyear research program to address the following relating to ﬁsh

eagle biology-

0 Characterization of the reproductive cycle of ﬁsh eagles at multiple sites

in Uganda.

0 Characterization of productivity and seasonality of ﬁsh eagle reproduction

at multiple sites in Uganda.

0 Characterization of the diet of ﬁsh eagles at multiple sites in Uganda.

0 Determine what happens to juvenile ﬁsh eagles between ﬂedging and

establishing an adult territory. Continue banding all eagles caught for

scientiﬁc research.

0 Quantify water quality indices at ﬁsh eagle study sites and assess the

associations between water quality and ﬁsh eagle reproductive parameters.

2. Assess the ability of ﬁsh eagles to act as biomonitors of other pollutants, such as

lead.

3. Develop banding techniques to assess whether a signiﬁcant proportion of the

Kampala marabou stork population is resident or migratory.

169 Resample the previously sampled avian and ﬁsh populations for mercury in ﬁve to seven years or more often if possible as per the sampling and analytical protocols of this research project. Particularly investigate the levels of 4,4'-DDE and mercury present in the Entebbe region of Lake Victoria.

Investigate the use of analytical techniques complementary to residue analysis such as stable radio-isotopic assays. Utilize this technique to answer specific questions relating to fish eagle and marabou stork biology e.g. examine trophic levels of prey of fish eagles at different sites to determine differences in diet composition.

Examine marabou stork nestlings in urban Kampala and a number of rural nests for differences in concentrations of heavy metals ﬁ'om exogenous and endogenous exposure routes.

Investigate alternative ﬁsh eagle capture techniques and reﬁne existing ones.

Techniques that exploit the ﬁsh eagle’s extreme territoriality require investigation.

Develop techniques for the capture of adult marabou storks.

Collect pathology specimens from dead birds for toxicological analysis

170 10. Establish appropriate collaborative links between ex-situ researchers and in-situ

individuals, private and government organizations.

11. Train appropriate Ugandan nationals in all sampling techniques and in

management of ﬁeldwork logistics and administration. Ensure all ﬁeld programs

are developed with realistic timeframes and adequate mandates for ﬁeld personnel

to achieve project objectives.

12. All projects developed should be multiyear research programs with a continuous

low-level in-situ presence by the research team.

13. Develop ﬂexible and realistic monitoring programs that can evolve with the

changing needs of Uganda, its people and environment. Ensure research

empowers Ugandans to make realistic and knowledgeable decisions regarding

their ﬁrture.

Serious consideration should be given to developing a battery of biomonitor species for assessing environmental change in aquatic ecosystems in Uganda. Recommended species are:

The African ﬁsh eagle (Haliaeetus vocifer)

The spot necked otter (Lutra maculicollis)

The Nile crocodile (Crocodylus niloticus)

171 Crocodiles have been used to assess the effects of persistent organic pollutants and heavy metals in Australia as have alligators (Alligator missispiensis) in Florida lakes (Jefree et a1 2001). Crocodiles however, are limited to certain waterways within Uganda. River otters (Lontra canadiensis) have been studied in relation to environmental PCB concentrations in North America (Wren 1991). The spot necked otter is common throughout most large waterways in Uganda. Both species meet many of the criteria of suitable biomonitors as stated in this thesis. The advantages of constructing multi- species biomonitoring programs in countries such as Uganda are many and varied and include:

1. maximization of economic, logistical, management and ﬁeld resources.

2. reduction of the effects of variation and nonpredictability between years in relation

to the timing of biological events within a species. This is achieved by allowing

sampling of at least one species to occur in all seasons.

3. increased accuracy of assessments by reducing the impact of confounding effects

that may occur in a single species monitoring program.

4. establishes a bank of indicator species which may be called upon to assess different

aspects of environmental change.

5. establishes a database and tissue bank speciﬁc for Aﬁican species.

172 As with any biomonitoring program, utilization of a multi species approach should occur in two phases: development of the biomonitor then utilization of the biomonitor.

173 References

Jefree R A, Markich SJ & Twining JR. 2001. Element concentrations in the ﬂesh and osteoderms of estuarine crocodiles (Crocodylus porosus) from the Alligator river regions. northern Australia: biotic and geographic effects. Arch. Environ. Contam. Toxicol. 40: 236—245.

Wren CD. 1991. Cause-effect linkages between chemicals and populations of mink and otter in the Great Lakes Basin. J. Toxicol. Environ. Health. 33:549-585.

Uganda Bureau of Statistics. 2002. Key Economic Indicators. 47th Issue. Ch 39. Uganda Bureau of Statistics, Entebbe, Uganda.

174